Q&A: Som Sinha, CEO, SAI
SAI’s solutions use our Visual Language Model (VLM) to blend computer vision with GenAI to turn traditionally ‘passive’ store surveillance systems into active, operational platforms for real-time store intelligence.
Can you tell us a bit about your background?
As CEO of SAI, I help retailers and brands drive store performance through active store intelligence, unlocking data tied up in passive systems and putting it to work to allow physical retail to operate more efficiently, profitably and safely, while enhancing customer experience (CX).
As a serial innovator, I bring 20 years of experience in tech development and business leadership across retail, mobile, and financial services, to take innovation from concept to multi-million-pound revenue-generating solutions. This ensures everything we engineer at SAI solves a true challenge for businesses rather than creating solutions looking for a problem, to deliver tangible and meaningful ROI for the retailers we work with.
What does your company do? What is your USP?
SAI’s solutions use our Visual Language Model (VLM) to blend computer vision with GenAI to turn traditionally ‘passive’ store surveillance systems into active, operational platforms for real-time store intelligence.
With real-time data from existing camera infrastructure, our AI-powered decisioning supports Loss Prevention (LP) and store safety while also unlocking operational efficiency gains. Offering a holistic store intelligence solution, we can open up context-rich customer insights beyond sales, inventory and superficial demand triggers, such as weather.
Wide-ranging data inputs, taken from real-time shopper behaviour in each store, are fed via our VLM technology into the solution’s GenAI engine. This builds precise, accurate and actionable intelligence, from shopper flow and retail media optimisation to heat mapping and behavioural analytics, for improved store performance.
Proprietary AI innovation within the platform analyses live data from video feeds using advanced VLM to build a comprehensive, contextual picture of how stores actually operate, surfacing meaningful operational signals, insight cues and timely staff alerts that drive estate-wide performance. Essentially, we help retailers run all their stores as their best stores.
What’s special about the platform and your approach? What advantage does it add?
Retailers are operating against a backdrop of rising costs across labour, operations and their supply chains, fragile consumer demand that puts increased pressure on pricing, and stubbornly high levels of theft and retail crime – complex and interdependent challenges which can no longer be effectively addressed through point solutions alone.
And that holistic approach to solving modern retail complexities and challenges is what differentiates our technology and sets us apart – we’re not trying to solve an isolated issue by engineering additional services onto our platform. Instead, any new challenge we start to address for a retail business works alongside the other issues we are already managing and solving, allowing us to take best practice and lessons from each implementation and drive them forward, applying new scenarios and innovation to the benefit of the retailer.
Our solutions also combine retailers’ current operations with intelligence to drive profitable improvements. For example, in the case of store operations, the solution evaluates available resources, such as colleagues’ workflows and availability on the shop floor, to determine the highest priority actions. Store staff then only receive alerts when new priorities take precedence over their current activities, streamlining operations while ensuring colleagues don’t become overwhelmed by alerts.
This not only immediately drives more productivity across the store, but the AI also starts to build a picture of whether labour models are sufficiently optimised to drive further operational improvements, enhance performance and increase revenue.
How are retailers using your systems to gain competitive advantage and what does best practice look like? Can you share a case study with us?
Last year, we worked with a leading UK frozen food retailer to deploy the industry’s first VLM-powered store intelligence platform, turning its existing CCTV infrastructure into a real-time operational intelligence system across approximately 1,000 stores.
Analysing live video feeds using our advanced Vision AI, the platform delivered contextual insights based on millions of in-store interactions gathered over 18months. By building a store intelligence platform – not another alert-generating technology – for each location, the retailer was able to interpret store activity as a whole, enhancing operations, improving customer experiences, reducing shrink and enhancing store safety.
With clearer insight into what is happening across individual shop floors and more widely across the estate, the retailer can now act proactively, rather than reactively, to create meaningful and profitable shifts in its day-to-day operations.
What challenges and opportunities do you see in UK retail for 2026? How will you address these challenges and turn them into successes?
As the ONS figures and data from the British Retail Consortium (BRC) remind us, retail crime – from shoplifting and organised retail crime to abuse of store staff – remains unacceptably and stubbornly high. And, sadly, that reality looks set to stay.
As inflation continues to rise – and with food prices tipped to be 50% higher in November than at the start of the cost-of-living crisis – consumers expect this financial strain to worsen retail crime rates. Our recent poll of 1,000 shoppers found that three quarters (76%) believe incidents of theft and abuse will rise in the next 12 months, and that’s on top of the 5.5 million incidents of shoplifting the BRC said were reported last year.
One of our priorities will be helping retailers enhance their crime prevention strategies, focusing on holistic approaches that move from reactive to proactive protection. While theft and abuse are rooted in criminality, they are also symptoms of broader operational blind spots and store complexity.
Stores need systems in place, not just to protect store workers when crime happens, but to proactively prevent instances of crime taking place in the first place. That requires store-wide data and real-time actions to effectively stop crime before it escalates, which is where we can help.
What is on the horizon for you as a company?
As retailers rethink the future of store safety and Loss Prevention, we are working with several large grocers on how to make Facial Recognition Technology (FRT) more transparent, accountable and privacy-first.
The technology has, to date, been met with scepticism from consumers – as we saw with the backlash against Asda’s FRT trial last year – and, arguably, for all the right reasons. In its current format, the technology has not yet reached a point where retailers can strike the right balance between effective protection, transparency and consumer trust.
Our research among shoppers shows that while there remains a disconnect between consumers and FRT, customers do understand the need for greater protective crime deterrents in-store – but acceptance remains low as trust barriers persist. In the same poll, however, shoppers indicated they would be more open to FRT use if thresholds and guardrails for privacy, ethics and transparent use could be better implemented and communicated.
And that’s something we’re eager to innovate around – and we believe there is a better way to use this innovation through rethinking how intelligent store technologies are designed and governed to keep stores safe whilst respecting shoppers’ data and privacy in a much more acceptable and ethical manner.
……………………….
To find out how SAI can help your retail operation, visit them online here SAI – Storewide Active Intelligence or connect with them here https://www.linkedin.com/company/sai-loss-prevention
