Picture the UK retail landscape in 2028. In the gleaming headquarters of tomorrow’s market leaders, AI agents orchestrate seamless operations across thousands of stores and online platforms. These systems predict consumer preferences before they’re expressed, optimise supply chains with unprecedented accuracy, and troubleshoot issues before they impact satisfaction. Customer loyalty soars, margins expand, and market share grows relentlessly.
Now picture the alternative future. Traditional retailers struggle with bloated inventories, frustrated customers, and shrinking relevance. Their executives watch helplessly as AI-powered competitors capture their most valuable customers. Board meetings become post-mortems on missed opportunities and the gap between leaders and laggards has become insurmountable.
UK retailers must now consider which side of history they want to be on.
The imperative for transformation
AI agents are no longer an experiment; they represent a fundamental shift in how AI can reason on a retailer’s data to transform forecasting, inventory management, customer engagement and, above all, business decision-making. According to Gartner, nearly 15% of daily operational decisions will be orchestrated by such agents by 2028. Yet, Gartner also predicts that 40% of agent-based AI projects could fail by 2027. Fears of a speculative bubble, industrialisation challenges, and budget constraints all fuel scepticism.
So, how do retailers bridge the gap between intent and aspiration, and actually see a clear impact from agent deployment? The answer to this question often lies in working backwards, identifying a clear and specific business opportunity while ensuring that strong data foundations are in place to enable AI agent systems to be truly impactful.
From hyper-personalised shopping experiences to autonomous supply networks – the possibilities are endless
The benefits for retailers looking to deploy AI agents are vast — both with internal and customer facing operations. Picture supply chains where AI agents not only forecast demand with pinpoint accuracy but also reroute shipments in real time to avoid delays, automatically balance stock across stores and warehouses, and flag potential waste before it happens. Instead of reacting to shortages or overstocks, these intelligent systems continuously learn from sales patterns, weather data, promotions, and even local events — transforming inventory management into a living, self-optimising network.
We are also going to see a shift where customers will increasingly rely on AI agent consultancy, moving away from direct interactions with brands. Domino’s “Voice of the Pizza” project to create a more seamless customer experience, is a clear example of this. By using AI and an interactive environment to fine-tune large language models (LLMs), the Domino’s team was able to parse through and gain insights from customer feedback on its subreddit page, resulting in gaining customer insights much faster and the ability to quickly identify areas for improvement.
Moreover, AI agents are set to radically transform customer service. A recent Capgemini study shows that AI agents could resolve up to 80% of customer inquiries in the first interaction, reducing customer wait times, creating a more frictionless experience and in turn, improving overall customer loyalty and ultimately, bottom-line profitability.
Start from the ground up — build the data foundations
With clear objectives to tackle in mind, retailers now need to think about whether their data foundations are fit for purpose. The saying that ‘AI is only as good as the data fed into it’ is undeniably true. Retail as an industry may benefit from a wealth of customer data, yet quality has been traditionally hard to measure, and therefore gauging agent success isn’t accurate. Furthermore, AI evolution, much like consumer expectations, is moving at a rapid pace. New models, tools, and techniques are regularly launched, and teams just can’t keep up. Underpinning all of this, however, is that many simply do not have the right, scalable data architecture in place to even think about rolling out AI agent systems in the first place.
Danone, the leading global food and beverage company, is an excellent example of a company that took the time to invest in a ‘future-ready’ data platform. The company is modernising its data infrastructure to a powerful platform built on a lakehouse foundation, which democratises access to analytics, is open source and has robust governance built in so that all the data is completely within its control. Alongside enhanced productivity and significant cost savings, the team expects this new platform to get high-quality, fully governed AI prototypes validated and into production at a much faster rate than previously possible.
With the right foundations in place, retailers can focus on building an AI agent system that is trained on their unique dataset to solve case specific challenges. This needs to go hand in hand with proper evaluation so agents aren’t judged on ‘gut feel’ alone, which is a recipe for inconsistent quality and costly mistakes. Data teams should also explore creating synthetic data that mirrors real customer information, helping agents learn quickly without the risks tied to live data. Take a grocery chain, for example. Synthetic shopper baskets could be used to train AI agents to recommend substitutions when items are out of stock, allowing realistic practice without damaging customer trust.
The time to act is now
No other sector is better poised to benefit from the rise of AI agents than retail. Many retailers have already made significant investments to better understand their customers and keep pace with rapidly changing behaviours, with many initiatives backed at board level to ensure a clear and consistent AI strategy.
The next step is clear; UK retailers that not only survive, but thrive, will be those that convert their business and customer data into actionable intelligence. Building strong data foundations and adopting fully governed AI agents will be essential for long-term commercial and operational success.
By Stefan Maczkowski, Retail & Consumer Industry Leader, Global, Databricks
