How ExpertVision Turns Store Cameras Into Anti-Theft Systems

OOMNIUM: Shoplifting costs the global retail industry over CHF 150 billion annually—and yet most retailers still rely on locks and security tags. Why has this problem remained unsolved to this day?

Alejandro Garcia: Because retailers have been stuck with two bad options: lock everything up, which kills sales and annoys honest shoppers, or put a guard on every aisle, which doesn't scale. Locks, tags, cabinets — none of that adapts to self-checkout or to how theft actually happens today. We have a solution that just runs on the cameras stores already have. That's the gap we stepped into.

OOMNIUM: Your “Virtual RFID” technology tracks products from the shelf to the checkout using standard cameras—without any facial recognition. How does this work in practice, and why is it a real breakthrough compared to what’s currently on the market?

Alejandro Garcia: Every shopper gets a virtual basket the moment they touch a protected item — no chip, no tag, nothing physical. Our cameras follow that product from the shelf all the way to checkout, and if it doesn't get scanned, we flag it. No facial recognition, ever — we're tracking products and behavior, not faces, which is exactly why we're GDPR and Swiss data law compliant. That's the breakthrough: RFID-level tracking with zero hardware on the product.

OOMNIUM: You’re a team with three very different backgrounds—AI and computer vision, retail operations, and platform strategy. Why does it take exactly this combination to solve this problem?

Alejandro Garcia: Our CTO David's spent 15 years in computer vision and machine learning — he's the reason our detection accuracy is over 95% with under 3% false alerts. Mate, our COO, worked with retailers at Coca Cola for over 20 years, so he understands how retailers work ensuring our solution fit their processes and deliver maximum value. And I've spent seven years building AI for retail and scaling a startup before this one. You need all three, or you either build something accurate nobody can use, or something usable that doesn't actually catch theft.

OOMNIUM: Within twelve months, you’ve completed technology validation, acquired paying customers, and launched a seed round. How have you used the capital so far—and what exactly will the current CHF 400,000 raise be used for?

Alejandro Garcia: We used our funds on building the core Virtual RFID™ detection engine to production-grade accuracy, and landing our first paying customers. We prepared and submitted our patent for VRFID and prepared ourselves for growth. The CHF 400,000 raise will fund the next stage of commercial scale-up: expanding the sales and customer success team, hardening the platform for multi-store rollouts, and funding the first steps into France and Spain.

OOMNIUM: You’re already in use at major Swiss retailers. What was the key selling point that convinced these customers—and what sets your sales approach apart from what other AI startups are trying in this market?

Implementation takes only a few days, and value generation can begin within a week.

Alejandro Garcia: We require minimal investment in new hardware and no integration with existing retailer systems. Implementation takes only a few days, and value generation can begin within a week. We have learned that the most significant benefit of our system is its case management capability, as many retailers still want to track incidents.

OOMNIUM: Your goal is to be present in 2,000 stores by 2030 and reach CHF 12 million in ARR. What are the two or three key drivers you’re focusing on right now to secure this growth trajectory?

Alejandro Garcia: We target retailers with more than 50 stores to ensure scalability after initial customer acquisition. Our pipeline includes 50 retailers and retail technology integrators covering over 20,000 stores. To scale quickly, we need to manage multiple new customer discussions and pilot programs simultaneously.