Retail is under constant margin pressure. Customer expectations are higher than ever. And the cost of mistakes – from empty shelves to mispriced SKUs – can be measured in millions.

SymphonyAI is demonstrating that its Connected Retail vision is resonating with retailers who are tired of firefighting and hungry for actionable insights at speed.

The tech company’s approach to connected retail deploys predictive and generative intelligence to convert real-time data into fast, actionable insights. By integrating customer data into everything from assortment planning to on-shelf availability, SymphonyAI is enabling retailers to view their entire business in one comprehensive view: what’s selling, what’s not, what’s on promotion, and even whether a shelf tag in Bournemouth is out of place.

“Connected Retail has been talked about for years,” says Jonathan Tye-Walker, an AI evangelist and strategic account presales specialist with SymphonyAI. “We’ve made it real by bringing together generative AI, predictive AI, and computer vision into an environment where category managers, store planners, and executives can all see the same picture – and act on it at speed.”

Beyond planograms: Automating the basics

SymphonyAI is building its presence in the Asia-Pacific market from its Singapore base with a focus on assortment and space solutions – an area that has long frustrated retailers. Traditional planograms are clunky, rule-based, and built in clusters. SymphonyAI replaces them with store-specific planograms created automatically, using AI-driven optimisation based on each store’s merchandising strategy.

That automation matters. “The key point is speed,” says Tye-Walker. “Tasks that once took teams hours or even days – painstakingly analysing data or physically walking aisles to check layouts – can now be completed in minutes.”

Linking category planning with SymphonyAI’s store intelligence solution gives managers real-time visibility into shelves, compliance, and stock levels. Computer vision scans shelves and feeds a dashboard with live conditions, workflow priorities, and AI-driven recommendations.

Cinde: The AI control centre

At the heart of SymphonyAI’s offering is Cinde – short for Connected Insights and Decision Engine.

“Think of Cinde as a control centre for analytics,” Tye-Walker explains. “You can set it up for all stores, all products, and within minutes you’re seeing dashboards that highlight sales, on-shelf availability, and customer behaviour.”

Cinde uses SymphonyAI’s in-house large language model, tuned specifically for retail, to provide diagnostic insights. “As a category manager, you don’t have to call on a team of analysts,” Tye-Walker says. “You ask Cinde a question – are my prices competitive, where am I losing sales, what’s trending – and it serves up answers, benchmarks, and recommended actions immediately.”

Cinde can analyse third-party price feeds alongside a retailer’s sales and stock data, flagging products that are out of line with pricing guidelines. It can model weather events, holidays, and store-specific demographics to recommend assortment changes. And it can even forecast the impact of delisting a SKU or shifting shelf facings – in one hypothetical example, showing a retailer that removing a $53 million snack line would result in a net loss of $13 million after transferable demand to rival products was factored in.

“You’re taking an ocean of data and boiling it down to a small pond,” Tye-Walker says. “And then telling the user where to fish.”

The bigger picture: From shelf to supply chain

While Cinde has quickly become SymphonyAI’s showpiece, the company’s offering stretches across the supply chain.

“We offer everything from warehouse management and replenishment to route forecasting and ERP-type functionality,” says Julian Miller, SymphonyAI’s global head of retail solutions success. “Our ultimate vision is to connect it all – to make retail data flow in a way that links planning, supply, pricing, and execution seamlessly.”

That matters because problems rarely exist in isolation. “If sales are down on a product, we can immediately see if it’s pricing, promotion, assortment – or something else entirely, like stock not making it to the shelf,” Miller explains. “It becomes holistic when you connect execution logic with supply chain visibility and in-store computer vision.”

Retail pain points: Cost, complexity, and customers

In conversations across Asia-Pacific, Tye-Walker says the same concerns keep coming up: “How do I optimise assortment and space? How do I manage promotions and pricing more effectively? How do I do more with fewer resources in an environment of rising costs and tighter margins? And how do I use AI without wasting money on generic solutions that don’t understand retail?”

For Miller, it boils down to three fundamentals: “Retailers are ultimately trying to drive more sales and volume, increase profit, and make operations in-store more efficient. Everything we build is aimed at those outcomes.”

One of SymphonyAI’s most prominent clients is the UK grocery retailer Co-op, which began working with the company in 2023.

Co-op’s challenge was that it controlled a mix of small urban stores and regional outlets and was grappling with wildly different customer behaviours. “People shopping in central London want a completely different offering to a tourist in Bournemouth,” Miller says. “On a sunny day, Bournemouth might sell out of ice cream by 10.30am.”

When SymphonyAI analysed the business, it found the right stock was not consistently reaching the right shelves in the right volumes. Co-op’s priority was to boost efficiency, ensure availability, and then convert that into sales and profit.

By using SymphonyAI’s assortment and space optimisation, Co-op began tackling categories one by one, resulting in strong sales growth and sharper on-shelf execution.

Real-time store intelligence

SymphonyAI’s system integrates into handheld devices used by store staff, offering real-time feedback. Using computer vision, it overlays shelves with colour-coded data: Green SKUs are correctly placed, blue items are in the wrong spot, red items don’t belong there at all, and yellow items are out of stock.

It can even trigger updates to electronic shelf labels, reflecting real-time price or availability changes. “This takes the manual gap-scanning process and automates it,” says Miller. “Store staff can focus on action instead of hunting for problems.”

The technology also calculates “missed on-shelf availability opportunities” – the sales lost from poor execution. In one hypothetical scenario, SymphonyAI flagged 1.8 million units of corn chip sales lost in a week due to out-of-stocks across a network. 

“That’s where it becomes real,” Tye-Walker says. “We can show retailers the exact cost of missed execution and the steps to fix it.”

The future of retail AI

While generative AI is capturing the headlines, Tye-Walker argues that retailers need retail-specific AI that understands the nuances of store operations and customer behaviour. “Some retailers are experimenting with generic AI tools, but they quickly hit limits. You need models built for retail – that’s what makes Cinde so powerful.”

For Miller, the ultimate goal is simple: “It’s about connecting the dots across retail. From shelf tags to supply chains, from pricing to promotions, we want to give retailers a single, AI-powered way to understand what’s happening – and how to act.

“This is about moving from drowning in data to acting on insight. From manual work to automation. And from reactive firefighting to proactive growth. That’s the transformation AI can deliver.”