Loops, a San Francisco-based company that builds AI search, shopping assistance and customer service tools for brands and retailers, has raised $3.5m in a seed round, the company announced today.
The round was backed by a16z speedrun, NTT Docomo, Connecticut Innovations, Alumni Ventures and Monday Capital. Loops said the money would fund product development, customer deployment capacity and international expansion.
The company combines search, product discovery, shopping assistance, customer operations, product enrichment and virtual try-on in a single platform, which it describes as an AI operating system for commerce.
Loops said AI is arriving in ecommerce one workflow at a time, with separate products for search, shopper questions, product descriptions and virtual try-on that retailers must stitch together.
Its platform is built around shared product knowledge, business context and customer journey data. According to the company, better catalogue data improves search, search gives the assistant context, and customer conversations show where product information is weak.
“The biggest mistake we see is treating AI as a collection of features,” said İlker Selim Zorluoğlu, president and co-founder of Loops. “A retailer does not need five different AI products that each understand one slice of the customer. It needs one system that understands the catalog, the business and what the shopper is trying to do.”
He added: “Commerce is where we are proving that model, and we think the same architecture can eventually become the AI operating layer for other industries as well.”
Shoppers using the platform describe what they want in plain language, the way they would to a friend or to ChatGPT, the company said.
Loops said its AI search then surfaces products without relying on keywords or filters. A built-in assistant answers questions and carries service conversations across connected channels.
Underneath, a product enrichment tool rewrites descriptions, attributes and catalogue data so products can be read by customers and AI systems. Virtual try-on lets shoppers see a product on themselves before buying.
Loops said the business began with a failed purchase. Zorluoğlu was in Bali with co-founder Ari Nazir when he tried to buy a gift for his two-year-old daughter, messaged the retailer and waited hours for a reply. He did not buy it.
The founders said they found shoppers were already telling AI what they wanted in plain language, while most of ecommerce still ran on keyword search, filters and separate support queues.
Nazir built the first prototype in a weekend, the company said. Loops now has more than 70 customers and pilots, has handled millions of shopper experiences and says 64 per cent of pilots have converted into paying customers.
Turkish supermarket chain A101 uses Loops to help customers navigate its catalogue and get assistance while buying. In a comparison Loops published, sessions involving the platform showed about 83 per cent higher observed purchase conversion than the comparison segment. The company said the result was observational rather than a randomised measure.
Fashion retailer Mavi uses Loops for AI search and shopping assistance, allowing customers to describe what they want by style, occasion or fit.
“AI is changing the interface between companies and their customers, but the lasting platforms will be the ones that can make it work reliably inside real businesses,” said Fareed Mosavat, investment partner at a16z speedrun. “Loops is starting in commerce, where the link between a better AI experience and business value is unusually clear. Its pilot conversion and early customer expansion show brands want a platform that can own more of that journey.”
Irish payments firm Trustap, which also builds for AI-led shopping, raised $10m in June to make marketplace listings transactable by AI shopping agents. Research from Coveo reported by Business Matters in September 2025 found 71 per cent of British consumers planned to use generative AI to help with their Christmas shopping.
Loops said commerce is the first of several industries it plans to enter, with financial services and energy next. The company believes the lasting opportunity in enterprise AI lies with systems built deeply inside individual industries rather than general-purpose models with a thin wrapper.
It is also building specialised small language model capabilities for tasks where, it said, domain knowledge and deployment constraints make a focused model the better tool.
