To put it simply, traditional search is dead. It has been for a while.

The search engine results page (SERP) we once knew has been completely rewritten. Gone is the era of users simply being shown a static list of ten blue links to trawl through. Today, search results are becoming more personalized and diverse; incorporating various media types and AI-generated overviews. With the rise of Large Language models (like ChatGPT, Perplexity or Gemini), search engines are evolving into “answer engines”, with users increasingly expecting direct answers, without the need for clicks.

From a user perspective this probably feels like an improvement, but for SEOs, marketers and brands, the implications are massive, with many unprepared for this AI-driven future. Traffic that was once coming to your site is being hijacked by AI, visibility is shrinking and attribution is more challenging than ever. What’s clear is the old SEO playbook is no longer working, and it’s urgently time for a revamp.

Why traditional SEO tactics are obsolete.

AI is simply the straw that broke the SEO camel’s back. But its legs were trembling for a while. For two decades marketers relied on the same old strategies aimed at gaming the system. We saw a rise in manipulative / spammy tactics like keyword stuffing, parasite SEO and content cloaking that resulted in the web being flooded by low-quality irrelevant content and poor overall user experience. 

However the algorithms got smarter. New anti-spam updates and the rise of AI-driven search means discovery is no longer about tricking Google with exact match keywords or link building, it’s about engineering content that is built for how modern search engines actually work. Google (for some time) has moved away from keywords and ranking, operating instead from vector embeddings and knowledge graphs. 

In other words: every piece of content, query, and concept is converted into a numerical “vector” in a vast, multi-dimensional space. The closer these vectors are, the more semantically related they are. That means Google prioritizes content that is contextually relevant, authoritative and genuinely helpful to users. 

At iPullRank, for years we’ve been talking about the need for a new evolution of SEO that operates within this new search paradigm. Something we call: Relevance Engineering. 

What is Relevance Engineering?

Relevance Engineering is multi-disciplinary approach that combines information retrieval (the science of how search works), AI (how machines understand and generate content), content strategy (how to create resonant content), user experience (how people interact with information) and digital PR (how authority and trust are built); with the goal of building a content ecosystem that aligns with both user intent and modern search engine expectations.

So what does this mean in practice?

• Content Engineering: you need to move beyond simple writing, to structuring content in clear and specific chunks that can be easily extracted and cited by AI. Every paragraph, every sentence, should be capable of standing alone as a relevant answer.

• Deep semantic understanding: look at the meaning behind queries, not just the keywords. This involves understanding “query fan-out” – how AI expands a single query into dozens of related questions – and ensuring your content addresses that broader semantic space. (We’ve even built a tool to help you do this).

• Build for citation, not just clicks: in an AI-first world, being cited in an AI Overview and AI Mode might be more valuable than a fleeting click if it establishes your brand as the authoritative source. Reevaluating old metrics will be key to your success.

• Use E-E-A-T as measurable signals: Expertise, Experience, Authoritativeness, and Trustworthiness are no longer abstract concepts; they are signals that Google’s AI models can assess, in part, through vectorized representations of authors, sites, and entities. Promote your experts, ensure your content is backed by authoritative sources, so the AI models have no choice but to cite you.

Traditional search is dead – and that’s a good thing.

The old SEO system was never built to scale with the modern internet. It incentivized shortcuts. It rewarded manipulation. And in the end, it made search worse for everyone.

In this new AI-driven era, gaining visibility is no longer about optimizing for ranking and success isn’t measured by traffic metrics. It’s about carefully engineering good-quality content to become the trusted source that AI models consistently reference and surface to your specific audience.

Relevance Engineering is an actionable strategy to not only stay ahead of the game, but drive more genuine leads to your website. Those that adapt to this shift in mindset will remain competitive, those that don’t, risk being left out of the search results altogether.

AI seems like it’s everywhere — doing everything from suggesting email subject lines to powering our smart homes. 

But has it reached its peak? 

Ask AI leaders like Sam Altman and Elon Musk and you’re likely to hear a firm “no”. Altman, in particular, has been vocal about his belief that AI will eventually surpass human intelligence. But what if we’re already seeing signs of the opposite? What if, instead of accelerating, AI is starting to plateau?

AI isn’t evolving on its own. It doesn’t learn like a human, there’s no gut-instinct, emotion, or lived experiences behind its development. Its capabilities are tied directly to the data that we give it. And when it comes to that data, even Altman and Musk could acknowledge that we’re beginning to hit a wall. 

So while AI may not have peaked yet, it might not be far off. 

Scraping the bottom of the web

Most of the growth we’ve seen in AI so far has come from feeding models huge amounts of data, scraped from articles, academic journals, websites, and social media platforms. But that supply is starting to dry up.

It’s what some experts are calling “Peak AI”. OpenAI’s co-founder has even compared the issue to fossil fuels — a finite resource that’s easy to exhaust, and impossible to replenish. 

And that’s where the issue lies. Without new data to train on, even the most sophisticated models will start to stagnate. And for businesses relying on AI to do more of the heavy lifting, that’s a real concern. 

When AI feeds itself

As new training data becomes scarce, a new risk is emerging. What happens when AI starts learning from its own output? This closed loop —where systems are trained on recycled or AI-generated data— can lead to a steady decline in performance, a scenario that is being referred to as “model collapse.”

For businesses that rely on AI in their workflows, this poses a serious threat. Model collapse can cause tools to produce inaccurate outputs — and in some instances, become entirely unreliable. 

The lesson is simple: if the quality of training data slips, so will the results. Garbage in, garbage out.

Why synthetic data can’t be a true replacement

To address the data shortage, many businesses are turning to synthetic alternatives, like AI-generated survey responses and simulated insights, designed to mimic real-world behaviours. 

But depending too heavily on synthetic data comes with its own risks. Without meaningful human input, there’s a danger that AI ends up falling back into a cycle of recycled, synthetic data, nudging us further toward model collapse. 

Over time, this can lead to repeated and amplified flaws or biases from older data, making each new iteration less accurate and more detached from reality. That’s a problem for any business trying to base decisions on those outputs. 

While AI may sound convincingly human, it doesn’t actually think like one. It draws from patterns it has seen before, meaning that synthetic data lacks the nuance that comes from real human insight. 

My advice for businesses? Used sparingly, synthetic data can help plug small gaps. But AI performs best when it’s rooted in reality. 

AI has reached a turning point, not a plateau 

So, has AI reached its peak? Not quite. But continued progress isn’t guaranteed. The growth we’ve seen so far has been driven by vast amounts of data, and it’s becoming clear that this momentum can’t be sustained.

What comes next is a turning point: a shift from quantity to quality. Businesses can’t rely on sheer volume of data or synthetic inputs to deliver results. Real-world insights, grounded in human experience, are what will keep AI useful and relevant. 

It’s not about having more data, it’s about having better data. 

What kinds of challenges are retailers facing today that make digital twins an appealing technology?

AV: Retailers today face a multifaceted challenge: meeting rising customer expectations, managing supply chain volatility, and maintaining operational efficiency — all while navigating growing pressure to reduce environmental impact. According to Coresight Research, 65% of brands and retailers struggle to manage their e-commerce visual merchandising operations, citing cost, emotional engagement, and consistency across channels as their top concerns.

Traditional approaches, such as in-store prototyping and high-cost photo shoots, are no longer sustainable. Digital twins offer a simulation-first alternative, enabling retailers to test and optimize experiences virtually before executing them physically. This not only reduces risk and expense but also accelerates speed to market. As Coresight notes, scalable and immersive content creation has become a top priority for retail CIOs — and digital twins are central to that shift. 

What does a digital twin look like in the retail space?

AV: ‘Digital twin’ is becoming a buzzword, but in retail, its meaning is highly specific and powerful. A retail digital twin might be a photorealistic 3D model of a product, a virtual store layout, or even a full shopper journey simulated with real-time data inputs.

Imagine a digital twin of a flagship store. A retailer could test 20 different shelf layouts. Rather than physically rearranging stores, they would model and evaluate each setup virtually, drawing on behavioral data to identify the most effective configuration. These are dynamic, data-driven systems that evolve as inventory, pricing, or shopper behavior shifts. So what starts as creating 3D digital versions of physical products ultimately becomes the building block for impactful AI-powered predictive tools that transform the entire retail experience.

3. How does this change (improve?) the customer experience?

AV: Digital twins shift the customer experience from static and reactive to dynamic and personalized. Instead of browsing generic layouts or static images, customers engage with immersive content far more closely tailored to their specific needs — from interactive 3D product displays online to AR experiences in-store. 

By simulating and optimizing the experience before launch, retailers can create online journeys that feel seamless and emotionally resonant. Coresight found that compelling visuals — like 360° CGI — not only increase consumer confidence in purchase decisions but also reduce returns and improve conversion. When a shopper can rotate a product, visualize it in context, or interact with it virtually, they’re more likely to stay, engage, and buy.

Where does Nfinite sit in this space? Where do you differentiate yourselves?

AV: Nfinite provides the infrastructure powering digital twins at scale. What sets us apart is our combination of visual fidelity, structured data, and enterprise scalability. 

We don’t just create beautiful 3D assets — we build simulation-ready content that integrates into AI-driven personalization engines and immersive commerce platforms. Our platform enables retailers to generate, manage, and deploy thousands of visuals — from product detail pages to virtual store environments — with the speed and efficiency traditional pipelines can’t match. 

That blend of quality, automation, and scalability is what allows our partners to move fast, and stay ahead. 

How is Nfinite helping major retailers leverage this digitally disruptive technology?

AV: We’re partnering with some of the world’s largest retailers including Lowe’s, Staples, and others,  to build full-scale 3D content ecosystems — not just for today’s needs, but for an AI-powered future.

It starts by digitizing their entire product catalog in 3D — thousands of SKUs rendered with precision and adaptability. From there, we enable automated content creation for omnichannel campaigns, tailoring visuals to different audiences, seasons, or contexts. 

Most importantly, we help integrate these digital assets into broader systems — powering product discovery engines, digital planning tools, and immersive experiences. This isn’t just about content creation. It’s about enabling a more intelligent, agile, and customer-centric retail model.

Let’s get one thing straight: content isn’t the problem. The lack of connection between content and revenue is.

Marketers are pumping more cash into content than ever before – and getting dangerously little back. 90% of marketing leaders have seen their content budgets balloon over the last five years. Yet only a shaky 39% feel confident linking that spend to actual revenue. The rest? Either praying no one asks, or holding up vanity metrics like they’re proof of pipeline. Spoiler: they’re not.

Welcome to the revenue gap – where killer content fails to make a killing, and marketing careers hang in the balance.

The data deluge is real – but so is the opportunity

We’re drowning in data. Every tap, scroll, and click generates a digital breadcrumb. Sounds like a goldmine, right? Except when 30% of marketing teams say they’ve lost customers due to bad data, and a third of their time is spent cleaning the mess up, you realise the gold’s been buried in rubbish.

Poor data not only wastes $16.5 million a year for enterprise firms – it tanks 26% of campaigns. And worse? It lets marketing output drift further from the revenue it’s supposed to drive.

That’s where AI comes in – not to patch holes but to plot a smarter course using better data. With the right tool, AI can be your compass in the chaos.

AI as your revenue co-pilot

AI and automation aren’t about making marketers obsolete. They’re about making marketers unstoppable. They find the important patterns in the data and show us what matters, so we can stop guessing and start making smarter decisions that lead to growth.

Platforms like Turtl show you, in real time, which content actually drives engagement, conversions, pipeline and, crucially, revenue. What’s resonating? What’s getting skipped? Where are we leaking attention? With Turtl, you can fix it now – not when you’ve already tanked half your budget on off-the-mark content.

We’re not talking shallow data that shows nothing. This is insight you can take to the CFO with total confidence.

Take predictive tools like Google Trends, or SEO heavyweights like Ahrefs that have built robust AI and automation capabilities into their platforms. They’re not just helping you create responsive strategies; they’re enabling you to get ahead of the curve for bigger impact. Couple that with behavioural analytics that reveal when your audience is most likely to engage, and you’ve got content that doesn’t just land – it converts.

Personalisation at scale = revenue at scale

A 2019 McKinsey study pegged the value of personalisation at up to $3 trillion. And yet here we are, still sending generic PDFs into the abyss.

With AI, you can tailor your content to thousands of unique buyer journeys, instantly. Platforms with built-in personalisation engines transform one-size-fits-all content into thousands of bespoke experiences. Not invasive. Not clunky. Just right.

This isn’t just noise. Real personalisation drives real results:

• $5M in pipeline influenced
• 4x more meetings
• 567% uplift in MQLs
• 1,500+ production hours saved (all from teams using Turtl, by the way.)

Optimise in real time, or get left behind

AI’s not here to admire your content. It’s here to test it, break it, and make it better.

Every piece of underperforming content is a missed revenue opportunity. Smart tools don’t just tell you something’s broken, they fix it. Layouts, visuals, timing, messaging – AI tests it all and suggests what to tweak next.

Take Turtl, for example. It gives marketers full visibility on drop-off points and engagement hotspots. If your CTA’s hiding in the dead zone, you’ll know – and our AI recommendations will show you how to fix it before your campaign flatlines.

Proof, not promises: reporting that stands up to scrutiny

Let’s be honest. We’ve all fluffed a marketing report or two. But in a world where CMOs are expected to deliver pipeline, “we think it worked” won’t cut it.

AI turns your raw data into clear, compelling dashboards that connect the dots between content and revenue. Tools like Tableau, HubSpot, and Turtl simplify the chaos, showing exactly how your content influenced pipeline, qualified leads, closed deals, and drove ROI.

Oh, and 96% of execs say this kind of reliable data would boost performance and productivity. You don’t say.

The takeaway: run revenue, don’t just report on it

The pressure is real. Tenures are shrinking. Budgets are ballooning. And the marketing leaders who can’t link content to revenue? They’re running out of rope.

But there’s hope, and it starts with better data, sharper insights, and AI and automation-powered solutions that help marketers make more impact with less heavy lifting. Because AI and automation aren’t just “nice to haves.” They’re your ticket to building a marketing machine that’s measurable, scalable, and revenue-generating by design.

Because the revenue gap isn’t a myth. It’s a monster. But with the right tech stack and the right mindset, you don’t just survive it.

You close it for good.

There’s plenty of discussion around AI strategies, but the real question is whether you can have an AI strategy without a solid data strategy? 

Setting up your data and AI strategy

Some argue that AI strategy should be built on a well-defined data strategy as data is needed to make AI work, while others see AI strategy as encompassing data needs within it. In fact, the more important sentiment is understanding that both need to be defined by your organisational goals. 

Whether it’s driving efficiency, enhancing decision-making, or freeing up resources for high-value work, you must ground your data and AI strategy in your goals and challenges, incorporating practical actions that deliver value to your organisation.

When you’re defining your data and AI strategy, using a data-driven framework can really help. 

At Transform, we recommend a top-down, bottom-up approach that teases out the practical and tangible actions that need to take place, keeping your goals and strategies in mind by asking what you’re trying to achieve.

Are you trying to attract new customers, deliver a better user experience, improve decision making etc?  

Your answers will more easily define what the bottom-up approach needs to achieve across the foundational levels, namely data and technology. You’ll then need to work on the enablers – people, process, systems and AI — and from there, you can narrow down what the required changes are that need to happen to deliver the desired benefits.  

This framework helps to identify and prioritise the right use-cases for tech, data and AI for value-driven outcomes.

It’s worth noting that, when you’re building your data and AI strategy, in addition to traditional data, people, process and technology components, you need to consider that outcomes need to be compliant and adhere to known regulatory and security requirements. 

The benefits of having a Data and AI strategy

A good data and AI strategy enables the effectiveness and efficiency gains promised by AI, such as:

• Making faster, better decisions: like when we helped Historical Royal Palaces write a digital and data strategy that allowed them to be bolder when bringing people to palaces and palaces to people.
   
• Using AI to do repeatable, mundane tasks, freeing up resource time to do more valuable work: like the work we did with DfE, helping to automate procurement processes for schools.  

Don’t forget to measure your success

The other component (often forgotten) is defining success and outlining the measurement framework for your data and AI strategy. What are you going to measure? How are you going to measure it? What limitations exist today and what new variables will you need to predict your success?

Defining what success looks like and establishing a measurement framework ensures that results aren’t just theoretical but tied to real gains. After all, you don’t want to miss the opportunity to tell your stakeholders that “this initiative saved X% time or Y£ or delivered Z% increase in engagement because our approach made us faster to serve” 

Everyone is talking about data and AI, but the real benefit is in the value they deliver for your people — making customer experiences better, being faster to serve, and being more efficient when it comes to operational process. 

Data readiness isn’t just about having data. It’s about making sure it serves a purpose. Without that clarity, an AI strategy is just an idea, not a driver of value.

It’s widely reported that the average person checks their phone 96 times a day, but let’s face it, that’s probably now a low estimate for a modern adult. This trend is not just reflective of screen dependence. It signals a continuous reveal of behavioural data: where you are, what you open, who you call, and even how long you linger on each app. Every moment of connectivity creates a digital footprint. 

For telecom operators, this stream of real-time data is an often untapped reservoir of insight. It reveals usage patterns, travel behaviour, content preferences, and signals of loyalty or churn. Used responsibly, this data can transform how telcos operate. Misused, it edges uncomfortably close to surveillance.

The rise of behaviour-led segmentation

Behavioural data can fuel smarter decisions, and that’s where its value lies. Modern operators are moving away from broad demographic segmentation toward behaviour-led models. Instead of seeing a customer simply as a 35-year-old urban professional, operators can now identify them as a weekend streamer, a weekday commuter, or a heavy international caller. This shift enables telcos to deliver timely, personalised offers such as data boosts on Fridays, international roaming passes before holidays, or entertainment bundles that reflect actual usage habits. Customers benefit from more relevant services, while operators unlock new revenue streams.

The same data can also help reduce churn, one of the industry’s most persistent challenges. By analysing subtle shifts, such as a drop in usage, a rise in complaints, or lagging service performance, operators can predict when a customer is likely to leave. They can intervene before it happens, offering personalised deals or improved support. It’s all about turning customer events into actionable insights and then deploying automated retention strategies in real time.

Walking the fine line between personalised and creepy

Yet, with all this power comes an uncomfortable question: how far is too far? At what point does personalisation become intrusion? Telcos sit at a critical crossroads, able to capture extraordinarily rich data but also responsible for protecting it. There is a clear ethical line between using behaviour to enhance a service and mining it in ways that compromise trust.

First and foremost, telcos must embrace data minimalism. Just because data is available doesn’t mean it should be collected or used without restraint. Operators should focus on metadata, such as call duration, time of day, data usage volume, and app categories accessed, which can legitimately inform service improvements and tailored offers. This type of information helps operators understand broad behavioural trends without infringing on personal privacy.

But there’s a clear ethical boundary when that metadata is used to infer deeply personal attributes, such as mental health status, financial hardship, or political views. For example, noticing an increase in late-night usage might inform the development of a time-based data plan. But using that same pattern to speculate on a customer’s emotional state is an overreach. The goal should be to enhance customer experience, not decode their private lives.

Transparency is also essential. Customers must understand what’s being collected and why. Clear, opt-in consent should be the norm, not the exception.

One of the best ways to maintain trust is to aggregate data before acting on it. Instead of targeting random users, operators can draw insights from broader groups, such as all commuters in a specific zone or a cohort of users with similar usage patterns. From this, they can still deliver individualised offers, but without the sense that someone is watching their every move.

The role of modern BSS in data responsibility

Modern business support systems (BSS) play a vital role here. Many legacy platforms lack the flexibility, speed, and visibility to manage data ethically and efficiently. BSS solutions that integrate real-time usage, apply AI-based segmentation, and automate offer deployment all within a secure, privacy-first framework are crucial. This ensures telcos can move quickly and intelligently without losing sight of customer trust.

The growing use of artificial intelligence raises the stakes. AI platforms can detect patterns far beyond human capability, predict churn with remarkable accuracy, offer opportunities in milliseconds, and segment audiences dynamically. But these capabilities must be balanced with explainability. If a customer receives an offer or is flagged as a churn risk, there should be a clear, auditable rationale behind that decision. 

AI should support, not obscure, the operator’s responsibility.

Applying an ethical filter: Helpful or invasive?

So, how can telcos draw the line between what is useful and what is unsettling? A helpful rule of thumb is this: would the customer perceive the action as a service or as a violation? Offering a data boost when usage spikes feels natural. Profiling a user based on app usage to infer sensitive traits, such as political views or immigration status, feels invasive. Responsible operators should run every data-driven interaction through this ethical filter.

As telcos evolve into digital-first, customer-centric providers, the question is no longer whether they can use behavioural data but how they use it and whether they can build trust in the process. Used wisely, data allows telcos to personalise offers, reduce churn, and deliver better value. Used recklessly, it risks eroding the very trust that underpins customer relationships.

The path forward lies in transparency, consent, and accountability. Telcos that embed these principles into their data strategy, supported by agile and ethical platforms, will gain a competitive edge and set the standard for what responsible connectivity should look like in the digital age. 

Behavioural insight can be a powerful tool for good, so long as it’s built on a foundation of trust. 

The UK’s National Health Service (NHS) had a brutal winter. An unseasonably bad flu season led to emergency rooms facing “exceptional pressure” as bad as the height of the COVID-19 pandemic, according to NHS bosses earlier this year. Clinicians find themselves working in situations where they are resource, staff, and (critically) time poor, with wait times growing unsustainably long as the health service struggles to handle over 1.7 million patient interactions per day.

One key pain point that medical professionals face is the manual documentation of patient discussions, with almost half of all GP time currently going towards administrative tasks. Artificial intelligence (AI) startup Tandem Health is aiming to change that with new tools that automate the documentation process, saving clinicians valuable hours that, they claim, can be better spent treating the public. We spoke to Ian Robertson, the UK & Ireland Director for Tandem Health, about his experiences as an NHS healthcare provider, Tandem’s AI solutions, and how they’re addressing issues ranging from AI hallucinations to ensuring confidential data protection and privacy. 

1.  Everyone knows the NHS is under pressure, but what does that actually feel like on the ground?

The pressure isn’t just a news story; it’s a daily reality for clinicians. Admin is a huge part of the problem. Every single consultation triggers a wave of documentation: notes, referrals, discharge summaries, coding. All of it is vital, but it eats up huge chunks of time. That trade-off — time spent on admin instead of with patients — is damaging. It limits access, pushes clinicians toward burnout, and affects the quality of care.

Our recent survey shows 56% of patients feel their doctor is too distracted by paperwork to give them their full attention. The data speaks volumes. There’s a clear need for tools that reduce the admin load and let clinicians focus on what matters most: patient care.

2. How much time does a typical GP spend on documentation?

Too much. For every hour spent with patients, GPs can spend nearly two hours on paperwork. Over the course of a year, that adds up to thousands of hours. Up to 40% of GP time now goes on admin. That’s time that could be used for decision-making, follow-ups or even just taking a break. It’s not just inefficient — it’s unsustainable. And it’s a major factor behind burnout and workforce attrition in the NHS.

3. What is Tandem Health building to address this?

We’ve developed an AI-powered medical scribe that listens during consultations and generates structured clinical notes in real time. It integrates with systems like EMIS, so documentation becomes seamless. But it’s not just about notes — Tandem can also produce referral letters and patient summaries, always under clinical supervision. Our goal is simple: give clinicians back their time so they can focus on care.

4. How does Tandem differ from off-the-shelf transcription tools?

Consumer voice tools aren’t built for healthcare. Tandem is. It understands clinical language, manages medical context, and fits into NHS workflows. It’s accurate, compliant and built with privacy at its core. That includes real-time processing, no audio storage and alignment with GDPR and NHS standards. We’re not just building tech — we’re building trust. That starts with understanding clinicians’ needs.

5. You’ve worked in the NHS yourself. How has that shaped the product?

Massively. I’ve been there, working long hours, dealing with relentless admin. I know what it takes for a tool to be genuinely helpful in a ten-minute appointment window. That’s why we build for the real world, not for labs. We don’t ask clinicians to change their way of working – we build solutions that adapt to them.

6. Hallucination is a known risk in AI. How do you manage that?

Clinical safety comes first. Tandem never replaces the clinician — it supports them. Every note can be reviewed and edited. We use domain-specific models, structured templates, and extensive validation. What the clinician sees is a safe, editable first draft that saves time and maintains control.

Our study with St Wulfstan’s confirms this. In that real-world setting, 95% of clinicians agreed Tandem’s notes accurately reflected the consultation. That kind of trust is essential.

7. What about patient privacy?

Tandem was built with privacy at its core. Audio is processed in real time and never stored. We meet NHS and GDPR requirements and are ISO 27001-compliant. We also don’t use clinical data to train models. With clinicians at the helm of our product development, patient confidentiality isn’t just a priority — it’s a responsibility.

8. What’s next for Tandem?

We’re expanding beyond general practice into outpatient departments and broader hospital settings. We’re also deepening integration with NHS infrastructure and supporting more roles across multidisciplinary teams. One of our biggest challenges is accommodating different workflows across organisations while keeping things safe and consistent. That’s why we’re investing heavily in interoperability, infrastructure and user experience.

9. Final thoughts on AI in healthcare?

AI can absolutely transform healthcare, but only if it solves real problems. Clinicians aren’t looking for novelty; they want relief. The best tools are the ones that give them time back, reduce stress and make care better.

And it’s already happening. At St Wulfstan’s, GPs using Tandem spent up to 68% less time interacting with the computer during consultations. Patients noticed too. The percentage who felt their GP was fully engaged jumped by more than 15%. That’s what progress looks like — not just better systems, but better conversations.

The adoption of AI in the workplace has been significant, sweeping through businesses at breakneck speed. Almost half (42%) are already embracing these powerful tools. Another 40% are actively experimenting. But alongside momentum comes with its challenges. As organisations deploy increasingly sophisticated AI systems, they also face heightened security risks and navigate uncertain regulatory ground; protecting both operations and human talent requires robust, forward-thinking safeguards.  

Equally as important to the success of AI is the operational foundation. Many organisations struggle with the absence of a clear AI roadmap, leaving them unable to progress beyond initial experimentation and ultimately fail to scale responsibly across teams. Without addressing this fundamental planning gap, organisations risk missing out on the transformative potential of AI to drive operational excellence, competitive differentiation, and sustainable growth. To truly harness AI’s potential – from driving efficiency to unlocking long-term growth – organisations must move beyond experimentation and invest in intentional planning. 

Realising AI’s full potential  

A survey conducted by Lucid Software revealed 49% of workers use it to automate repetitive tasks — freeing them to focus on higher-value work instead. Workers also recognise AI’s broader potential. Some cited improved productivity (62%), as well as seamless integration with existing workflows (41%), cost savings through consolidated tools (40%), and enhanced communication and decision-making (38%) as key potential benefits of AI adoption. 

Yet, despite decision-making being a top advantage, only 23% of workers currently use AI for this purpose. Bridging this gap will require a thoughtful, inclusive approach — aligning AI with business objectives and continuously refining its role to maximise its impact.   

A divide in perspectives  

While there’s broad optimism about AI’s potential, the enthusiasm varies across organisational levels. For instance, 68% of executives believe AI will enhance their job satisfaction. However, this drops to 53% among managers and is only 37% among entry-level employees. This disparity highlights a critical challenge. If organisations want to successfully implement AI, they must bridge this perception gap and demonstrate its value to employees at all levels. 

Many workers are already using AI for basic tasks, but its full potential remains untapped. Only 26% use AI for synthesising ideas or research, and just 19% leverage it for designing diagrams. This suggests that while AI adoption is growing, organisations have yet to integrate it in ways that drive meaningful innovation.  

The key to AI’s effectiveness lies in its intentional integration. Organisations must align AI with existing workflows to enhance productivity without creating friction. A common misconception about implementing AI is that it’s only useful if it produces perfect results. However, that mindset overlooks its true value. 

Right now, AI isn’t ready to replace entire workflows. It’s most effective when augmenting specific tasks, removing bottlenecks, and enabling teams to focus on higher-value work. Organisations that recognise and embrace this incremental approach will see the greatest impact. 

Tackling challenges head-on  

While 88% of companies are implementing AI guidelines to protect their operations and employees, communication around these efforts is lacking, leading to confusion and misalignment. For example, only 29% of entry-level employees feel confident their company actually has these rules in place. Combined with concerns around job security (33%), this has resulted in a third of businesses reporting a resistance to change as a top challenge when implementing AI.  

As AI continues to evolve, the need for ongoing education and training becomes increasingly critical. 

Executives are more likely to seek independent learning opportunities, 39% compared to 13% for entry-level workers. This underscores the need for an intentional, accessible, and continuous AI education framework for all employees. Effective change management strategies that communicate AI’s benefits, address concerns empathetically, and involve employees in the transition can build trust and demonstrate that AI complements rather than replaces human effort.  

The journey to success  

Workplace attitudes towards AI are mixed, ranging from enthusiasm to unease. Despite AI’s ability to enhance productivity and decision making, these advantages are often overshadowed by anxiety, resistance, and lack of understanding. 

To address these challenges, leadership must implement deliberate strategies to create organisational alignment, provide comprehensive support systems, and deliver targeted training on AI utilisation. By cultivating collective understanding and equipping team members with appropriate resources, companies can maximise the transformative benefits of AI. 

For those unfamiliar, what is Freshworks and how does it differentiate itself in the crowded AI space?

At Freshworks, we build AI-powered software that makes IT and customer support teams more efficient and effective.

Over 73,000 companies choose us over larger competitors like ServiceNow and Salesforce because we offer enterprise-grade alternatives that are incredibly easy to use, implement and scale. We are the antidote to bloated, complex service software.

In this crowded AI space, many companies are tapping into the same foundational LLMs. The difference is what you build on top of them and how fast your customers can get value.

At Freshworks, our AI isn’t just a chatbot or a bolt on. We’ve built a connected system of AI teammates (Copilot, Agent and Insights), deeply integrated into our platform, trained for practical CX and EX use cases, and designed to deliver value from day one.

Our differentiation comes down to four things: 

• Uncomplicated by design, easy to implement, adopt and see results
• Rapid impact, customers get measurable ROI fast, often in weeks, not months
• Purpose-built for service, our AI is customized for customer support and IT
• Secure and responsible – with trusted partners such as Microsoft, Amazon, OpenAI, Anthropic, Meta and data companies such as Snowflake and Databricks, we build AI capabilities that are  safe, trusted, reliable and grounded in context. 

We don’t just drop an LLM into your system. We fine-tune it with domain expertise and build it into workflows that actually help your teams scale.

You’re announcing Freddy AI Agent Studio – what is it, and what sets it apart from other Agentic AI product suites?

We’re unveiling the next evolution of the Freddy Agentic AI Platform—designed to make it even easier to reap the work productivity benefits of Agentic AI. With no-code agents that can be created and deployed in just minutes, we’re removing the delays and complexity that hold teams back on platforms such as Salesforce and ServiceNow. At the center is Freddy AI Agent Studio, a no-code platform that lets teams build custom AI agents to automate customer service tasks.

Why this matters: Customer service teams across industries such as retail, travel, financial services, manufacturing, and SaaS can now quickly deploy AI to handle high-impact tasks such as flight rescheduling, loan authorization, and customer verification – without needing more technical resources. This speeds up support, reduces costs, and ensures scalability even in lean environments.

We’re also rolling out four more updates across the Freddy Agentic AI Platform: 1) Email AI Agents that learn and automate ticket resolution with no human intervention needed, 2) AI Insights that identify and surface up IT issues before they escalate, 3) Unified Search AI Agents that can help find answers instantly across business applications, and additional capabilities on AI Copilot that helps teams work smarter and faster.

Freddy AI Agents are already used by over 1,600 Freshdesk customers. Now they can be deployed across the business in just five minutes, while Salesforce and ServiceNow products require months or even years of costly deployments before agents can get up and running.

We’re giving every business the power to deploy their own customer support AI agents in five minutes – not five months. No code, no complexity. Just real outcomes, fast.

What are some of the new capabilities being introduced across the Freddy AI platform?

Our AI Agent Studio is a game-changer. Picture a retail support team heading into the busy holiday season. They need help managing a flood of “Where’s my order?” questions. With AI Agent Studio, they can build and launch an AI Agent that connects to their order system and handles these queries automatically – all without a single line of code. In just minutes, the AI Agent  is live, taking automated actions to track orders, update customers, and free up human agents for more complex issues.

Within the AI Agent Studio customers get access to:

• Skills Library – pre-built templates of skills required by AI Agents to take actions in commonly used applications including Shopify and Stripe
• Skills Builder – a visual, no-code environment to design and deploy custom skills for AI agents to autonomously resolve service requests like processing a return

Freddy AI Agents can deflect up to 70% of incoming tickets and go live in under five minutes. Business users can build and deploy AI Agents without need for any developer or technical resources. 

How does this rollout compare to what we’re seeing from legacy players like Salesforce and ServiceNow?

Competitors require months of costly and laborious implementation. With Freddy, you drag, drop and launch. A customer can go from idea to automation before the workday ends.

Freddy AI Agents are live in minutes, not five months, unlike Salesforce and ServiceNow—who offer promises of low-code but still take weeks or months to get agents live—Freddy AI delivers real automation in under five minutes. That’s not a pilot. That’s production-ready, now. We uncomplicate work so customers can focus on results, not red tape.

Can you share some real-world examples of how customers are using Freddy AI today?

Customers are seeing real impact across every layer of the Freddy Agentic AI Platform.

Hobbycraft automated 30% of support requests with Freddy AI Agent, freeing agents and boosting customer satisfaction by 25%. Bergzeit reduced translation work by 75% with Freddy AI Copilot, processing 200,000+ tickets. And Five9 uses Freddy AI Insights to identify and close service gaps before they impact customers.

Over 5,000 companies now use Freddy AI products, seeing up to 70% ticket deflection and 50% productivity gains. Freddy AI is a force multiplier for teams.

What are the most common use cases you’re seeing across industries?

Companies across Retail, Travel, Financial Services, Manufacturing, Tech, and more will benefit from our new Agentic capabilities. They span a wide range of use cases across industries like: Order tracking and management; flight booking management; payments, bill sharing, and subscription management; and inventory management.

Our AI agents can take action on these tasks end-to-end, without human intervention.

 What kind of ROI or productivity gains are customers seeing with Freddy AI?

The numbers speak for themselves. Freddy AI Agents are deflecting up to 70% of incoming tickets. Copilot is delivering up to 50% productivity gains. Bergzeit auto-triaged over 200,000 tickets and reduced translation workload by 75%. That’s not just efficiency – it’s transformation.

How does Freshworks approach pricing for these new AI capabilities?

Our customers told us they’re tired of the confusing pricing and hidden fees they experience at competitors. So we made Freddy AI Agents a simple, flexible, “pay as you go” model. The new AI Agent Studio is currently in “early access” so there’s no fee to try it.

How is Freshworks staying ahead of the curve in AI-driven CX?

We’re not chasing AI hype – we’re building practical solutions that deliver real outcomes. Our platform is cloud-agnostic and model-neutral, drawing from over 40 LLMs including partnerships with Microsoft OpenAI and AWS. This flexibility enables us to adapt more quickly, optimize for performance, and consistently select the best tool for each task.

What’s next for Freddy AI and Freshworks’ approach to agentic AI?

We’re focused on continuing to deliver usable, efficient, and high-impact AI that drives real value. Customers choose us because they don’t have time or budget for complex deployments. They want solutions that work out of the box, are cost-effective, and drive productivity – which is exactly what we deliver. That’s how we’ve earned defections from legacy players like ServiceNow and Salesforce, and why we’ll keep winning.

Another day, another scandal over generative AI trained on stolen data. This morning, social media giant Reddit launched legal action against artificial intelligence startup Anthropic, claiming the company’s AI assistant was trained on Reddit users’ data. It’s the latest in a long, long, long line of ethical and legal pitfalls lining the technology’s path to assumed eventual profitability. AI luminaries (and also tech industry lobbyist and one-time politician Nick Clegg) are even going so far as saying that AI companies won’t be profitable or competitive if they have to pay for the data they need to train their models. ChatGPT-designer OpenAI openly admitted to the UK Parliament that its business model couldn’t succeed without stealing intellectual property and data.

“It would be impossible to train today’s leading AI models without using copyrighted materials,” the company wrote in testimony submitted to the House of Lords. “Limiting training data to public domain books and drawings created more than a century ago might yield an interesting experiment, but would not provide AI systems that meet the needs of today’s citizens.”

James Evans is the head of AI and engagement products at Amplitude. Previously, he was the Co-founder and CEO of Command AI, which was acquired by Amplitude in October 2024. We caught up with him to get his take on the AI data privacy issue, as well as the future of personalisation, and walking the thin line between a better customer experience and an intrusive one. 

1. AI is a profoundly data-hungry technology. How do you think organisations can balance AI’s insatiable demand for private, sometimes copyrighted data with the need to respect privacy?  

I believe organisations need to flip the traditional approach on its head. Don’t design AI products or services and then frantically scramble to find the data you need to power them. Instead, start with the data you know you can use legally, and then build from there. Sometimes this means being less ambitious about your AI initiatives, but it ensures you’re on solid ethical ground from the beginning.

Also, I’m a strong advocate for letting users choose. Be transparent by saying, “Hey, if you want to use this functionality, you need to give us more information about you.” My experience is that when the benefit is clear and tangible, users are often much more comfortable sharing their data. It’s about creating that value exchange that people can understand and opt into.

2. A lot of the sanctity of privacy and copyright laws were quite flagrantly ignored to build the large AI models in the first place. As companies like OpenAI try to build the next generation of models, do you think they’ll continue to take the same approach, or can the industry’s relationship towards stolen data be rehabilitated? 

I think OpenAI and other model companies recognise that if we delete the incentive to produce good human-generated content, we will end up in a place with worse AI technology. Social media and journalism is a good cautionary tale – we saw the incentive for good journalism go away when everyone was consuming stuff on Facebook et al instead of generating ad dollars for publications. Then you saw a new economic model develop: subscriptions. I already see a lot of conversation around new economic models emerging to reward people for creating good content that AI then leverages. 

3. From a CX perspective, what’s your take on the increasingly frontloaded presence of AI tools in everything from search bars to word processing apps? Is it actually making the customer experience better? 

AI in customer-facing applications is moving beyond superficial implementations toward more meaningful integration. Language-based interfaces are emerging as standard entry points for complex applications, enabling more intuitive user interactions that drive efficiency. There is a shift away from flashy, standalone features toward embedding AI into core functionality where it can deliver tangible value.

Multi-modal AI capabilities are particularly transformative for user assistance, analysing not just text but broader session data and user behavior to provide deeper insights and more accurate recommendations. This enables smarter and more personalised interactions with customers, helping solve long-standing user experience challenges such as reducing navigation complexity, minimising search frustration, automating repetitive tasks, and providing contextually relevant suggestions based on actual usage patterns rather than predefined pathways.

However, success depends on moving beyond gimmicks to focus on real utility. Companies that can deliver this while maintaining appropriate privacy controls and data governance will be best positioned to improve customer experiences meaningfully.

I think it’s worth emphasising that we are all getting much better at prompting AI. In fact, I think many users – especially those from groups who aren’t super fluent with software interfaces – are better at prompting AI than they are navigating link trees and dashboards. I think as that trend continues, people will expect and breath a sigh of relief when they see a text input in an app, instead of a complicated interface. But undoubtedly interfaces will still exist for high subtlety or creative work.

4. What are the consequences for companies that get this balance between intrusion and personalisation wrong? 

Not getting the balance wrong between personalisation and intrusion can have serious business consequences. For example, when companies bombard users with poorly timed, irrelevant popups and notifications, they create “digital fatigue” – users begin to automatically dismiss guidance without even reading it. Most traditional popups are closed immediately, meaning users are reflexively dismissing them before even processing the content.

Excessive or poorly targeted intrusions erode trust, increase bounce rates, and damage both conversion and retention metrics. We’ve seen cases where overly aggressive in-app messaging actually decreased feature adoption because users began avoiding areas where popups frequently appeared.

Conversely, companies that strike the right balance see dramatically different outcomes. By using behavioural data to deliver personalised guidance precisely when users need it – not when the company wants to promote something – organisations can achieve drive engagement and adoption.

The key is using AI-powered targeting and “annoyance monitoring” to ensure guidance appears at moments of maximum relevance. This means tracking not just if users engage with guidance, but actively differentiating between normal closures and “rage closes” (when users immediately dismiss content), which signal poor timing or targeting. Companies that implement these more sophisticated, user-respectful approaches maintain trust while still delivering the personalised experiences that drive business outcomes.

5. What’s on the horizon for the conversation about AI, personalisation, privacy, and the user experience? 

I believe we’re going to see several significant shifts in the AI landscape. First, enterprise applications will move away from bolting on AI as a separate feature and instead truly embed it into core functionality. We’ll see AI capabilities woven into workflows in ways that feel natural rather than forced or gimmicky.

I also expect the AI ecosystem to become much more diverse. Companies will adopt a multi-provider approach rather than betting everything on a single large language model. This shift recognises that different AI models have different strengths, and organisations will become more sophisticated about choosing the right tool for specific contexts.

One particularly exciting development will be the rise of specialised AI models that demonstrate superior performance in specific domains. These purpose-built models will often outperform general models in their areas of expertise, creating opportunities for startups to carve out valuable niches.

Multi-modal AI capabilities will transform how we approach user assistance and analytics. By processing not just text but images, user behaviour, and other data streams simultaneously, these systems will enable much deeper insights and more accurate recommendations than we’ve seen before.

All of this technological advancement creates tremendous opportunities for both startups and enterprises to address long-standing user experience challenges through smarter, more personalised interactions—while hopefully maintaining appropriate privacy safeguards. The most successful organisations will be those that balance innovation with respect for user boundaries.

8. How does the launch of DeepSeek in January (along with the promise of other AI models developed outside of Silicon Valley) change the industry’s prospects? 

I think the emergence of models like DeepSeek is awesome for two reasons.

First, it clearly demonstrates that there’s a ton of innovation out there that intelligence—not just money—can unlock. There’s significant room for smart people to make an impact in this space – it’s not just about hurling dollars at bigger GPU farms. That’s incredibly exciting because it means we don’t have to rely solely on Moore’s Law type scaling to get better performance. We can achieve breakthroughs through clever engineering and novel approaches.

Second, it serves as a wake-up call that China can seriously compete in AI. Our leaders should assume that China will be very competitive in this space, and that Western countries won’t enjoy some type of durable intellectual advantage. This reality should inform both business strategy and policy discussions around AI development and governance.

8. Given that the Trump administration is currently working very hard to ensure that the US regulatory landscape won’t exist (or at least be very different in a few short years, or months), what does this mean for AI companies who were, almost to a one, being sued and/or investigated for unethical and illegal use of private information?

It’s really hard to say with certainty how this will play out. The regulatory landscape for AI is still evolving globally, not just in the US. That said, I do appreciate the administration’s emphasis on enabling startups to innovate and not anoint incumbents as the only players allowed to do interesting things. There’s a genuine risk in over-regulating emerging technologies that you end up simply entrenching the position of companies that are large enough to navigate complex compliance requirements.

At the same time, we shouldn’t mistake regulatory flexibility for a complete absence of accountability. Regardless of the formal regulatory environment, companies still face reputational risks, potential consumer backlash, and market pressures that can meaningfully shape behaviour. Plus, many AI companies operate globally and will still need to address standards set in places like the EU.

I believe the industry itself will need to develop better self-governance approaches. The companies that proactively build ethical data practices and respect privacy boundaries will be better positioned for sustainable growth, regardless of short-term regulatory changes.

Recent discussions by EU institutions made it abundantly clear that deploying artificial intelligence (AI) in justice and home affairs is no small feat. Despite its transformative potential, AI’s adoption comes with significant hurdles, such as data quality, infrastructure readiness, and ethical compliance, which are just the tip of the iceberg. These challenges resonate across industries, but their impact is particularly acute in sectors where public trust, safety, and governance are non-negotiable.

At a recent roundtable hosted by eu-LISA, the European Union Agency for the Operational Management of Large-Scale IT Systems in the Area of Freedom, Security, and Justice Industry, discussions underscored a contradiction in AI adoption. While the technology promises to enhance efficiency and decision-making, its use in operations can expose foundational weaknesses in readiness that range from integration barriers to ethical dilemmas. Only when these gaps are addressed will AI deliver on its potential.

The Challenges: Insights from the Roundtable 

Several recurring themes emerged during the eu-LISA roundtable, including infrastructure gaps, data and compliance, ethical complexities, and talent shortages. While many of these are known, it is important for us to relook at how they are impacting public institutions. 

Infrastructure Gaps

Many public institutions are underprepared to scale AI from experimentation to full deployment. As highlighted by the European Commission and echoed in the Nutanix Enterprise Cloud Index (ECI), integration with existing systems remains the number one challenge when scaling AI workloads.

Data and Compliance 

Quality, security, and the accessibility of data are ongoing challenges and high-risk sectors like justice and home affairs are especially vulnerable to gaps in data governance, which undermine AI’s reliability. Compounding this is the stringent compliance required under frameworks like the EU AI Act.

Ethical Complexities 

Public sector AI applications often intersect with sensitive domains like biometric data and predictive policing, where transparency and fairness are paramount. As the roundtable participants noted, for society to trust AI, these systems must be practical and ethically sound.

Talent Shortages

Both the roundtable and the ECI findings point to a lack of skilled personnel as a bottleneck. Over half of organisations recognise the need for additional training and recruitment of the right people to support future AI initiatives.

Infrastructure as a Launchpad for AI

AI is only as effective as the environment it operates in. During Nutanix’ session, “Slow In, Fast Out (with AI),” we talked about how infrastructure is like the foundation of a house. If it’s shaky, nothing you build on top will last. Public institutions cannot afford to deploy AI systems on shaky foundations. Whether it’s predictive analytics or generative AI, scalable platforms are critical for ensuring seamless operations.

A robust Enterprise AI platform is essential for simplifying deployment while maintaining flexibility. By leveraging Kubernetes, these platforms can enable hybrid and multicloud environments to handle workloads with agility. For public institutions and private enterprises, adopting a “start small, validate use cases, and gradually scale” approach helps reduce risk while maximising return on investment.

Building Trust Through Governance

The EU AI Act provides a framework for balancing innovation with societal safeguards. However, compliance is just the beginning. At the roundtable, eu-LISA emphasised the need for independent testing and monitoring mechanisms to build trust in AI systems. These safeguards ensure that high-stakes applications, like biometric identification, meet stringent transparency, safety, and accountability standards.

Organisations must also invest in model governance to address the lifecycle of AI systems. Centralised repositories for AI models and robust access controls and monitoring tools can mitigate risks while ensuring compliance with evolving regulations. This is another area where Enterprise AI Platforms play a critical role. 

Collaboration and Human Expertise

One of the biggest takeaways from the roundtable was that no single organisation can solve these challenges alone. AI in justice and home affairs demands collaboration across government, industry, and academia. It’s not just about sharing technology; it’s about sharing perspectives, experiences, and solutions.

And let’s not forget the human side. While AI can streamline decisions and processes, it’s the people behind those systems who ensure everything stays aligned, ethically and operationally. In support of this, the ECI report reveals that over 50% of organisations are investing in training programs to upskill their teams. This democratisation of AI knowledge fosters a culture of innovation and resilience.

Turning Challenges into Opportunities

The discussions at the roundtable echoed a sentiment we see often: the challenges associated with the technology aren’t going away. But they’re also not insurmountable. Generative AI, for example, is reshaping priorities, particularly around security and privacy. This shift drives organisations to modernise infrastructure, rethink compliance, and invest in their workforce.

By addressing these challenges head-on, institutions can turn obstacles into stepping stones. Taking a strategic approach, one that balances technical readiness with human-centric governance lays the groundwork for AI systems that don’t just work but truly make a difference.

Within the UK food and drink manufacturing sector, there’s a £14 billion growth opportunity waiting to be unlocked. The details were revealed in a new report: Future Factory: Supercharging digital innovation in food and drink manufacturing. 

The report explains how the implementation of AI, automation, and digital technologies are key to seizing this untapped potential. Leveraged properly, they can lead to accelerated productivity gains throughout the sector.

The importance of AI has been further compounded by the Government’s AI Opportunities Action Plan unveiled in January. It outlines how AI can help to “turbocharge” growth and boost productivity.

The value of AI and machine learning is clear. Therefore, if we take the UK’s food and drink manufacturing sector as an example, how does AI and ML work? More importantly, what’s standing in the way of progress?

AI applications in manufacturing

Hidden inside the plant and machinery of every factory in the world there is a wealth of data. Once unlocked, this data can help to improve the overall equipment efficiency (OEE).

AI and machine learning, alongside deep domain expertise, are key to liberating and contextualising this data.

Half of the world’s top 12 food and beverage manufacturing companies – including names like Muller, Mars, ADM, Weetabix, Hovis and Diageo – are working with IntelliAM to harness the transformative power of their data. 

We work by installing sensors that harvest millions of data points within a variety of supply chain components, the data is contextualised into a wide range of categories such as speed, pressure, product, flow and lubrication timing. This is then overlaid with reliability data indicating why faults occur. 

These faults and problems can range from issues with vibration and oil condition to temperature of induction motors and loading of Programmable Logic Controllers (PLCs).

Once we have the knowledge of these factors, we equip the sensors with effective alarms, allowing for the health and efficiency of equipment to be monitored. This forms an individual stamp for each component that highlights crucial information such as finding the root causes for errors or mitigating future process shortfalls which, in turn, increases productivity.

For one of our clients, we implemented an OEE analysis and predictive maintenance system which harvests 400 million data points per month. This discovered consequential data that enabled us to predict future stoppages – through this non-invasive method we were able to increase their line performance by 6%.

Exploring the barriers to AI and ML adoption

At present, the top manufacturers are only accessing around 1% of their potential data.  

For long enough, there have been hurdles in the industry which have limited production leaders from shifting their mindset be open to these new, transformative systems.

Yet while the Future Factory report states that 75% of the food and drink industry values the benefits of digital technologies, it also explores how they are held back by several cited obstacles.

These perceived barriers include the ability to instantly prove return on investment, negative preconceptions of AI and how to integrate it into legacy systems and equipment, as well as a significant skills gap, and rigid food and safety procedures.

But what if these perceived obstacles are more imagined than actual barriers? Mental roadblocks rather than real-world challenges?

Food and drink manufacturing is caught in a vicious cycle. Financial pressures restrict technology investment, leading to a stagnation in productivity, which, in turn, limits further capital investment. 

But manufacturers don’t need to rebuild factories or invest in brand-new equipment. The answers lie within their existing assets.

Integrating AI and ML into the existing food production process

Machine learning that integrates with existing assets – no matter the make or age of the machine – means companies don’t need big capital investment to achieve the first steps to converge with advanced technology.

Another highly voiced concern in connection to AI is around job displacement. However, AI and ML work most effectively when they are coupled with domain expertise. A knowledgeable, well-trained workforce will always be needed in order to deliver impactful results. 

AI and machine learning need teams of engineers to tag, code, and instruct the system so it can learn the algorithm to become self-sufficient. AI is therefore contributing to creating talented, skilled workforces.

It’s also important to address another misconception within food and drink manufacturing industry. Many believe that to get ahead of the curve and be a part of the AI and machine learning movement they need to abandon legacy systems and replace them with brand-new expensive machinery. This is a major misconception.

There are millions of data points hidden inside existing plant and machinery. They just need the right tools and technologies to liberate and, most importantly, contextualise them.

Having access to in-depth data insights helps to drive more informed decision-making, too. Manufacturers have the power of foresight – anticipating and fixing problems before they occur and determining training requirements.

Seizing the AI and ML opportunity

The challenges outlined in the report aren’t as difficult as they appear.

Data can be extracted from all machinery – regardless of the model, brand, or age. 

Factory floors can continue business as usual whilst asset data is gathered in the background. This data can then be used to bridge productivity gaps and drive manufacturing forward.

This is more important than ever given that global food demand is always increasing to support population growth. Over the next 25 years, we’ll need to produce more food than humanity has ever produced before. This means food manufacturers will need to embrace technology and innovation to help meet demand.

Ultimately, whether manufacturers are ready or not, technology convergence is coming. AI and ML are redefining what’s possible in the food manufactuirng sector.

With continued uncertainty in the market about global economic conditions and the pressure to control supply-chain costs, there’s more need than ever in 2025 for newer, smarter operational strategies. As we edge further into this year, it’s important for businesses to consider how they can continue to drive greater efficiencies and lower costs, while still evolving to modernise their tech stack and prepare the business to pursue new opportunities for growth. 

As AI continues to redefine how businesses compete and operate, Agentic AI has emerged as an especially promising solution for a more intelligent and self-sufficient way of working. In a shift from assisted intelligence to genuine autonomy, industry experts anticipate accelerating interest in agentic AI investment, predicting enterprise adoption to spike to 33% by 2028 — an exponential leap from less than 1% in 2024. 

Yet it’s not only about Agents, and realising the desired outcomes from AI requires a slightly broader strategic perspective. With the right blend of AI patterns and an accessible, intuitive no-code platform, these intelligent AI-powered tools can empower organisations to unlock unprecedented levels of productivity, fostering a collaborative ecosystem where human and digital talent work in sync to drive innovation. 

Breaking down the AI triad: Generative, predictive, and agentic 

While AI provides an extremely broad spectrum of transformative capability, it can be distilled down into three essential patterns – generative, predictive, and agentic AI – which each serve distinct purposes. Gen AI takes patterns learned from vast datasets and uses them to generate novel content — from text and images to music and code. Predictive AI, on the other hand, analyses historical data to forecast future outcomes, providing crucial insights for informed decision-making across various business functions. Unlike the former two, which are largely passive in their operation, agentic AI is capable of thinking and acting autonomously based on learned behaviours. It can perform complex tasks, automate workflows, and adapt to changing conditions with minimal human intervention.

As one of the latest developments in artificial intelligence, agentic AI operates with a high degree of autonomy, while maintaining real-time adaptability and human oversight. It analyses data, understands contexts, and executes complex actions within pre-defined parameters. Powered by machine learning, large language models (LLMs), and reasoning engines, it continuously applies and acts upon its intelligence while working alongside human employees.

Agentic AI and the workforce

The powerful capabilities of Agents can immediately create concerns about loss of jobs; this theme dominates many news cycles these days. However, we believe this actually creates an opportunity for most information workers to create new value and allow job expansion. For the individual, Agentic AI reduces the time spent on routine activities, such as data entry, synchronising information across systems, or completing highly repetitive tasks. This creates space for employees to focus on more strategic, creative and high-priority tasks. This shift doesn’t replace human roles—it co-exists with them, ensuring people work in harmony with AI for greater efficiency, creativity, and decision-making.

Furthermore, the use of new AI agents is rapidly requiring the building of many new skills and talents. In terms of job creation, this shift is already taking place across various industries. According to a 2025 Job Market Research report, AI-related job postings peaked at 16,000 in October 2024, showing rapid growth in newly established roles. AI’s integration into day-to-day operational processes necessitates new roles in developing, deploying, and managing these intelligent systems. 

This need for rare human talent subsequently creates knowledge gaps for companies fighting to maintain competitiveness in the tech ‘space race’. As a result, the demand for tools that make AI initiatives more accessible for a broader range of employees has soared. Businesses who empower employees at all levels to work alongside AI create a more agile, adaptable, and collaborative workforce.

Agentic AI on the front line 

Insiders predict that Agentic AI will be one of the biggest strategic trends over the next few years. Gartner predicts that by 2028, one-third of interactions with GenAI will invoke autonomous agents to complete tasks. Across every industry, businesses are beginning to apply Agents to optimise processes, improve productivity, and unlock new revenue streams. With the power to ‘learn on the job’ and gradually improve over time, agentic AI is particularly well-suited for supporting staff in stakeholder interactions. 

Timing is everything — especially when it comes to effectively managing the workforce. While basic AI-powered chatbots allowed companies to shift customer services from limited hours to 24/7 support, agentic AI takes this a step further, making interactions more dynamic and context-aware. 

Retailers, for instance, can use agentic AI to answer customer queries, process refunds, or make product recommendations, reducing the need for human agents to handle routine tasks. Unlike traditional automation, these AI-driven agents learn from each interaction, improving their responses over time. When escalations do occur, agentic AI analyses them to refine its approach and ensure human agents receive the most relevant context before stepping in. 

This human-digital collaboration is where the true potential of AI lies. Rather than replacing jobs, agentic AI enables employees to focus on solving complex customer issues, fostering stronger relationships, and delivering a superior experience.

Getting started with no-code AI building

Agentic AI is becoming a prevalent tool for business transformation. But with the growing concerns regarding the scarcity of tech talent, organisations are left wondering where to begin implementing agentic AI. 

To address this problem, experts predict a drastic increase in demand for citizen platforms that provide simpler tools and which unify diverse AI stacks and seamlessly orchestrate machine learning, generative AI, and agentic automation. As such, no-code platforms are emerging as an important solution, rapidly gaining popularity due to the shortage of developer skills.

Taking a less technical approach to software development, no-code platforms can be the ideal entry point for agentic AI implementation and deployment. These platforms enable employees to build applications with no programming skills required. This allows for the easy customisation of intelligent agents and support portals, while eliminating the daunting complexity of traditional coding — saving both time and money, and bridging knowledge gaps. 

As we progress into 2025, it’s up to organisations to find ways to implement this technology beneficial for both the workforce and the bottom line. It all boils down to strategic planning, resourceful upskilling, and responsible AI agent implementation. The future of work is AI-augmented, not AI-replaced. The key to success lies in human and digital talent working together, empowering businesses to scale AI innovation while at the same time realising operational efficiencies.

AI is revolutionising industries, but it comes with its own set of challenges. Navigating the evolving landscape of AI can be complex, with rapid technology updates and legal changes. As a result, some companies are uncertain about adopting AI and concerned about how to approach it. Others fear being left behind and feel pressured to act quickly.

However, rushing into adopting AI without planning use cases, and assessing potential hazards, is risky.

The Hidden Risks of AI

From bias and discrimination to privacy and security concerns, and lack of transparency, AI requires careful risk management. This is especially true for sectors like healthcare, finance, or transportation, where the impact of failures can be severe.  In addition, AI tools are now more accessible to the public. These tools can produce very convincing content, which may not be accurate or good quality. 

Responsible AI, combined with a clear AI strategy, is crucial to address these challenges. It takes a holistic approach, tackling social, ethical, compliance, and governance risks for organisations. 

Organisations must have a robust AI Governance framework in place, including policies and risk management processes. These measures ensure that Responsible AI principles are effectively implemented, and supported by the necessary structure. It’s also crucial that they align with the company’s AI strategy, values, and goals.

Building a Strong Governance Framework

AI Governance should tie in with existing company governance structures and programmes. Aligning with international standards, such as ISO 42001, ensures that key elements of AI risk management are covered. Another important step is employee training in the benefits and risks of AI. This builds awareness in the organisation to increase effectiveness and reduce risks. In addition, it complies with The EU AI Act AI literacy requirement to train employees using or building AI systems. Together, these measures increase transparency, define accountability, and mitigate risks in business operations.  

It’s essential to understand the unique AI challenges for each company and the sector in which it’s based. For example, in healthcare, it is critical to make sure patient privacy, quality of care, and data security are protected. Responsible AI policies need to be tailored to these, to make sure they are adequate and effective for the company. This bespoke approach is essential to develop guidelines and governance that work in practice.

Keeping Up with AI Laws

Staying ahead of legal changes in the AI world is vital. Global updates on AI laws and regulations are now released at a similar pace to technical news on the latest models. Companies need to make sure their AI strategies and policies are aligned with the latest legal developments. This is especially important when working across several regions, with differing legal obligations. A proactive approach is essential to navigate this changing landscape and ensure compliance. This is key in safeguarding the company’s reputation and legal standing. 

A Catalyst for Innovation

When implemented correctly, AI can deliver positive benefits for organisations.

Project SEEKER is one example of this. It was developed by BJSS, in collaboration with Heathrow Airport, Microsoft, UK Border Force, and Smiths Detection. The AI system automatically detects illegal wildlife in luggage and cargo at borders. This alerts enforcement agencies. The project has aided in the fight against illegal wildlife trafficking with over 70% accuracy.

AI Governance plays a key part in project success and can be a powerful driver of business innovation and growth. It provides a secure and compliant environment for AI adoption and development.

The Future of AI

Addressing bias, privacy, and regulatory standards means companies can mitigate legal and reputational risks.Responsible AI is more crucial than ever. AI is now being used in many different contexts, and tools are more widely accessible to the public. Companies must carefully assess use cases and manage risks to make the most of the technology. Responsible practices, clear AI governance, and regulatory compliance are vital for sustainable success with AI. By focusing on these, businesses can ensure that AI continues to benefit both their operations and society at large.

There is no doubt that sales have rapidly evolved and changed in the digital age, with many sales leaders feeling that traditional cold outreach falls short in today’s competitive business world. This has called for a rise in automation, which relies on data for its success. Data is the key to creating a bridge between impersonal, ineffective outdated outreach and meaningful and successful sales conversations. 

Many in sales roles use a “spray and pray” tactic, hoping contacting enough people with a standard message will lead to success. However, this tactic is unsurprisingly declining, as more customers expect personalisation and relevance to them with sales offers. Decision makers are bombarded with generic calls and emails that fail to address their unique business challenges. Today, buyers expect relevance, industry-specific insights, and solutions tailored to their organisation. Automation has become essential for scaling outreach, but its success hinges on data. By using data to segment industries, target specific roles and personalise messaging with insights into business goals or pain points, sales teams can shift from impersonal mass outreach to valuable conversations which resonate with B2B prospects.

Data is Key to Modern Sales

Data is the key differentiator in identifying, segmenting and targeting prospects. There are several types of data which drive automated sales, including; demographic data which focuses on characteristics at the individual level, such as job title, seniority, location, and professional background. 

B2B Sales Data 

In B2B sales, this data is crucial for identifying decision makers or influences within an organisation. For example, a SaaS company targeting mid-sized companies might focus on IT directors or CTOs in specific industries. 

Defining an Ideal Customer Profile using demographic data allows teams to narrow their focus to prospects who are most likely to convert, ensuring that outreach efforts are spent on the right people. This data also enables targeted messaging, for instance, emphasising technical capabilities when reaching out to IT leaders versus ROI when targeting CFOs. 

Behavioural Data 

Another major asset is behavioural data which provides insights into how prospects engage with your brand across various channels. This includes website visits, email opens, link clicks, webinar attendance, or even interactions with your competitors. Behavioural signals can indicate a prospect’s level of interest and readiness to engage, helping sales teams prioritise leads more effectively. For example, if a prospect repeatedly visits a product comparison page or downloads a whitepaper, automation tools can flag them as “hot leads” and trigger or encourage personalised follow-ups. Behavioral data not only improves lead scoring but also informs outreach timing, engaging prospects when they’re most active increases the likelihood of a response.

Firmographic Data 

Firmographic data describes the many different attributes of a business, such as industry, company size, revenue, geographic reach, and growth trajectory. For B2B sales, this is one of the most critical data types because it ensures outreach aligns with the broader needs and goals of the target organisation. 

For example, a marketing agency might use firmographic data to make an appropriate pitch to a small startup versus a multinational enterprise, tailoring its solutions to align with its unique challenges and budgets. Firmographic data also enables account-based marketing strategies, where highly targeted campaigns focus on specific high value companies or accounts.

Intent-Based Data 

Intent-based data is a kind of purchase data signal that helps understand the exact signal based on buyer signals, focused on the type of cookies across third-party sites to create intent-based information to target those actively buying vs wasting energy for less active buyers. These can include enriched data from website visitors to understand why visitors from websites are not converting and have proactive engagement with these. 

By combining these data types, sales teams can automate personalised outreach that feels human, is highly relevant to the prospect’s needs, and builds a strong foundation for conversion. 

Bringing a Human Touch to Automation

Automation doesn’t need to be removed from the human touch either, especially when it’s powered by data. Automation enables sales teams to deliver highly personalised communication at scale, making outreach more relevant and engaging. 

By using data to understand a prospect’s role, industry, and specific challenges, automated systems can craft messages that resonate on a personal level, even in high-volume campaigns. For instance, an automated campaign targeting UK-based retail companies might reference seasonal trends or recent industry developments, leading to significantly higher response rates compared to generic messaging. 

Personalisation driven by automation doesn’t replace the human touch, it amplifies it, allowing sales teams to focus their time on building genuine connections with prospects who are already engaged.

The Future of Data and Automation in Sales

The future of sales automation lies in the increasing integration of advanced technologies like AI-driven insights and predictive analytics. These tools enable sales teams to predict behaviours, identify high potential leads, and personalise outreach with greater accuracy. For example, predictive analytics can highlight which accounts are likely to convert based on historical patterns, while AI can craft tailored messaging that aligns with a prospect’s industry or challenges. However, as data usage becomes more sophisticated, the need for ethical practices and transparency grows equally critical. Businesses must prioritise compliance with regulations such as GDPR and ensure their outreach respects privacy and fosters trust. Staying ahead in this evolving landscape requires organisations to treat data strategy as a living framework, regularly updated, refined, and aligned with technological advancements, new laws and ethical standards.

Data has proven itself a transformative force in sales, turning cold outreach into warm, meaningful engagements through personalisation, prioritisation, and precision. Sales professionals who embrace data-driven automation while maintaining the human element are in the best position to thrive. The most successful sales strategies combine the power of technology with a commitment to building trust and genuine connections at scale. While tools and data play a vital role, sales success remains fundamentally about understanding people and delivering value in ways that resonate.

Besnik Vrellaku is the CEO and founder behind Salesflow.io, a leading force in Go-To-Market (GTM) software revolutionising B2B lead generation for SME’s using multi-channel sales technology and supporting over 10,000 users with modern prospecting solutions used by the likes of Hubspot, Hibob and Gocardless. 

A new report by Nexthink warns that a lack of readiness to adopt AI could undermine organisations’ efforts to adopt the technology. 

Organisations will spend $5.61 trillion on IT in 2025, with $644 billion going towards Generative AI alone. According to a new report from digital employee experience (DEX) management company Nexthink, 66% of IT decision makers say their organisation rolls out a new application, tool, or platform every month.

Despite widespread enthusiasm for the technology among companies looking to create efficiencies, cut costs, and replace human workers (in both the public and private sectors — even in the US government), Nexthink’s report warns that a “lack of employee readiness to adopt and confidently use AI could see investments go up in smoke.”

Nexthink: the Science of Productivity

Nexthink’s report, ‘The Science of Productivity: AI, Adoption, And Employee Experience’ report details the findings of a survey of 1,100 global IT decision makers. In the report, 95% of IT leaders said that they expect the upcoming wave of AI-powered digital transformation to be the most impactful and intensive seen thus far, as the latest phase (agentic AI) promises better, more independent AI solutions that can act with less human supervision. 

However, the majority of IT leaders (92%) surveyed also said they believe this new era of digital transformation will increase digital friction. The abiding opinion was that fewer than half of employees (47%) have the requisite digital dexterity to adapt to technological changes. Almost nine-in-ten leaders said they expect workers to be “daunted” by new technologies like Generative AI.

“Organisations are spending trillions on IT to digitally transform, but without their people on board, it’s a fast track to failure,” said Vedant Sampath, CTO at Nexthink. “Too many employees are left grappling with unfamiliar AI tools because they lack digital dexterity: the ability to confidently embrace new technologies. IT teams, meanwhile, are flying blind without visibility into where things are going wrong. Transformation isn’t just about rolling out new tech; it’s about enabling people to use it effectively. If businesses don’t end this digital dexterity crisis, they’ll end up with cutting-edge AI tools – but a workforce that can’t use them. That’s a one-way ticket to watching AI investments go up in smoke.”

The risk of laying GenAI failure at employees’ feet

IT leaders agree that resolving this digital friction and improving the employee experience must be a priority. The risk, they say, is that failed AI adoptions eat up budgets without creating tangible value for the business. 

At the same time, 42% of IT leaders admitted to Nexthink that they struggle to put an exact monetary value on AI investments, while 93% want to improve their ability to identify underperforming investments.

Regardless, IT leaders still anticipate a 43% rise in the volume of AI applications over the next three years. 

The data matches up with a report by the World Economic Forum, which found earlier this year that 41% of employers intend to downsize their workforce as AI automates certain tasks. 

But this rapid expansion of AI adoption is, Nexthink says, stretching IT teams to breaking point. Almost 70% admitted that there are too many users in the organisation for IT to provide adequate adoption support for everyone. Without proper guidance, application rollouts suffer, leading to lower productivity (61%), reduced collaboration (51%), increased IT support tickets (46%), and higher employee dissatisfaction (46%).

“Digital transformation lives and dies by the employee experience,” added Sampath. “If IT teams can’t effectively guide employees through adoption, businesses will never unlock the full value of their investments. DEX is no longer a nice-to-have; it’s business critical. Without it, IT leaders will struggle to measure impact, let alone maximise returns, and risk seeing their transformation efforts stall before they even get off the ground.”

As generative AI continues to evolve, we’re beginning to see the next generation come to life: Agentic AI. Traditional AI is designed to answer a single prompt. By contrast, Agentic AI can perform multi-step tasks and work with different systems to achieve a more complex goal. 

Customer service is a good example of an Agentic AI use case. An AI agent might handle inquiries, respond to support tickets, take follow-up actions, and even escalate complex issues to human agents. This ability to automate entire workflows and make decisions across systems is what sets Agentic AI apart. Deployed correctly, it could be a game-changer for many industries.

The promise of Agentic AI is immense. Gartner forecasts that by 2028, a colossal 15% of all day-to-day decisions will be made autonomously by AI agents. 

AI agents can drive efficiency, cut costs, and free up IT teams for strategic work. However, deploying them also presents its share of challenges. Before deploying Agentic AI, businesses must address issues that could compromise the reliability and security of these systems.

1. Enhancing model reasoning and insight

As the name suggests, Agentic AI systems use multiple interacting agents to make decisions. One agent might function as a “planner” to set a course of action, while others act as “critical thinkers” that assess and adjust these actions in real-time. This creates a feedback loop where each agent continuously improves its decision-making ability.

But for these systems to be effective, the underlying models need to be trained on realistic, high-quality data — data that reflects the complexities of the real world. This requires continuous iterations, sometimes involving thousands of scenarios, before the model can reliably make critical decisions.

2. Ensuring reliability and predictability

With traditional software, we provide explicit instructions — step-by-step code that tells the system exactly what to do. Agentic AI, however, relies on a more autonomous approach, where the AI decides the steps needed to reach a desired outcome. While this autonomy offers efficiency and scalability, it also introduces unpredictability, as an agent might take a less predictable path to the solution.

This isn’t a brand new phenomenon. We saw a similar situation with the early versions of LLM-based generative AI like ChatGPT. Back then, outcomes were occasionally random or inconsistent. In the past couple of years, however, quality control initiatives like human feedback loops have made these systems more reliable. 

The same level of investment will be necessary to reduce the unpredictability of Agentic AI. The technology can’t be useful unless it can be trusted to take reliable action. 

3. Protecting data privacy and security

Privacy and security considerations  are paramount for the organisations considering Agentic AI. 

Since AI agents often interact with multiple systems and databases, they’re likely to have access to sensitive data. Similarly to Generative AI where every piece of data provided to the model gets embedded within the system, Agentic AI could inadvertently expose a business to vulnerabilities, such as data leaks or malicious injections.

To address these concerns, companies can start by isolating data and implementing robust segmentation protocols. Additionally, anonymising sensitive information, such as removing personally identifiable data (like names or addresses), before sending it to the model is key. For example, a financial institution using agentic AI to process customer requests should ensure that transaction details are anonymised to prevent exposure of sensitive data.

At a top level, right now, Agentic AI can be categorised into three types based on its security implications:

• Consumer Agentic AI: These models interact directly with end-users, so security measures are crucial to prevent unauthorised data access
• Employee Agentic AI: Developed for internal company use, these systems carry less risk but can still expose sensitive information to unauthorised employees. For instance, companies might create their own GPT-like system for internal tasks, but it needs safeguards to protect confidential data
• Customer-facing Agentic AI: These systems serve external clients and must be designed to protect both customer data and proprietary business information

4. Ensuring data quality and relevance

For agentic AI to perform at its potential, it needs to be able to draw on accurate, relevant, timely data. Many AI models struggle to deliver that pipeline because they don’t have access to real-time, high-quality data — whether that’s an issue with the data itself, or the pipeline that supplies it.

A Data Streaming Platform (DSP) can address these challenges, allowing businesses to collect, process, and transmit data in real-time from multiple sources. For instance, developers can use Apache Kafka and Kafka Connect to integrate data from various sources, while Apache Flink facilitates communication between different models. 

Agentic AI systems can only succeed, avoid errors, and generate accurate responses if they are built on trustworthy, up-to-date data.

5. Balancing ROI with talent investment

Deploying Agentic AI requires considerable upfront investment, not just in hardware and infrastructure, but also in acquiring specialised talent. Companies may need to invest in memory management systems, new GPUs, and new data infrastructures, while in-house teams must be trained to build inference models and manage AI systems.

Although the initial return on investment (ROI) is reliant upon a careful, methodical implementation, the long-term benefits can be significant. In fact, tools like Copilot are already being used to autonomously write and test code, showcasing that businesses can start integrating these systems today.

Despite its challenges, Agentic AI is poised to revolutionise business. With the power to outpace Generative AI, it’ll drive decisions at scale across industries — from healthcare to autonomous vehicles. 

Though the path to adoption may be tough, the impact will be massive, reshaping how businesses operate. The key? Investing in quality data, solid security, and the right infrastructure. Once in place, Agentic AI can unlock huge efficiencies, help decision-making, and fuel growth.

Many enterprise leaders who are bullish about GenAI hold the view that data cleansing and architecting must come before the technology’s rollout. But is this missing the bigger picture?

Data inputs impact analytic models. That still rings true in some cases. However, the emergence of unstructured data processing, whether via Large Language Models (LLMs) or traditional regression techniques, offers immediate opportunities that don’t require the complete overhaul of existing systems. Companies I speak to with GenAI success stories don’t have flawless data lakes or necessarily cutting-edge analytic stacks. Instead, they’re finding ways to move fast and unlock value with imperfect data environments. So, what’s their secret?

Not all use cases are equal

Some organisations are reporting huge efficiency gains and cost reductions from using GenAI while others are seeing modest ROI. More often than not, this comes down to use case selection. This is no surprise. It’s been a defining element of success in analytics for years.  

The greatest challenge in the analytics process is widely viewed as this initial phase, translating business challenges into use cases. How might data analytics be used to optimise your inventory? How can data help streamline tax credits? Could you improve your customer service by being more personalised?

Currently, many organisations base their selection of GenAI use cases on risk profile. This is just one of the key factors for GenAI’s success. Use cases must align with the LLM techniques that we know to perform well. This means picking use cases that really leverage the amazing capabilities of what an LLM can do and staying away from those where LLMs will fall short. 

The chatbot wave

While chatbots dominate GenAI applications due to customer service and process automation, their real value extends far beyond simple conversation. LLMs can be used to scan the news and summarise information to provide alerts. For example, you could input the cities and dates individuals at a company are traveling and create automated alerts sharing potential disruptions picked up on the internet scans. While an investment firm could use an LLM to sift through the news each day and provide succinct summaries for key news that could be used by analysts to assess against its portfolio. These are just two low-risk use cases where LLMs tend to perform well, summarising large amounts of unstructured data and providing succinct or even structured outputs that can be easily used.

Additionally, the use cases described require little data from the companies building the automation, send very little data externally, and can provide references to where the information came from so that the user can validate the sources. This is perfect for companies to ‘dip their toes’ into GenAI and serves as a great ramp to the technology with minimal risks.

Converting unstructured data into structured data

While many associate GenAI with chatbot solutions, others are finding that leveraging LLMs to convert large amounts of unstructured data into structured tables of data can prove impactful. Imagine using an LLM to scour the websites of your competitors to pull all their pricing into tables of data, which are organised in rows and columns (e.g. name of competitor, product description, current price). This leverages the magic of this new technology in a use case that most organisations would view as both safe and requiring minimal dependency on the quality of their internal data.

The challenge then becomes, how do you guide the organisation to the right use cases to start with? The answer lies in internal culture and education.

Change management

Successful GenAI adoption goes beyond merely putting the right technology into more hands. Organisations must  provide education and foster an environment that embraces these new techniques. The concepts are not difficult, and learning how to apply the technology to a myriad of domains is within reach with the right mentors guiding the team.

Change management has been a longstanding requirement for organisations to achieve analytics maturity. Whether helping the organisation learn to leverage self-service data wrangling and modelling tools or applying Machine Learning (ML) techniques to problems. However, in the context of GenAI, change management becomes less of a “nice to have” and more of a non-negotiable necessity for success.

Education is critical. Companies deploying analytics tools often accompany this with one-off training. However, the most successful organisations blend practical skills (which includes the training to get them there) with foundational knowledge. Take data visualisation. While teams need to know which buttons to press, they also need to understand the principles underpinning effective visual communication. This combination of “how” and “why” creates far more impactful results than technical step-by-step guides. The same principle applies to GenAI. Organisations should have a systematic approach to bringing people on the journey using education and training, not just technology. 

This can be summed up in fostering an AI literacy culture. And with this, there must also be guidance on when it’s appropriate to use the technology. GenAI can and will provide new capabilities, but not all problems are GenAI problems. It could be ML, automation, visualisations and other techniques. Organisations that understand this are far more likely to get the most out of GenAI technology.

Final thoughts

Flawless data, data readiness, and underlying infrastructure isn’t a prerequisite to GenAI success. What matters most is how organisations prepare and support their people through the transformation that the technology entails.

The good news? Critical success factors of education, knowledge sharing and change management are within the control of enterprise leaders. Companies don’t need to wait for perfect conditions to begin their GenAI journey. They can start today by building the right foundation of skills and understanding, confident in the knowledge that technology adoption is a gradual process. 

Savvy organisations recognise that humans, not technical perfection, will determine whether their GenAI initiatives excel or falter. By investing in people’s ability to understand and leverage new tools effectively, they’re setting themselves up for success.