Stoyan Ivanov spent two decades inside Coca-Cola watching the same experiment fail. A brand would build a dedicated app for its retail partners, roll it out across a market, and wait. Eighteen months later, adoption would hover somewhere around 15%.
The retailers, including kiosk owners, corner shops, and independent distributors who collectively move more than $5 trillion in goods globally, would keep ordering the way they always have: by phone, by voice note, by text.
Ivanov, who ran Coca-Cola’s e-commerce and digital ventures across Europe before leaving to start his own company, nFuse, thinks the industry drew the wrong lesson. The problem was never that retailers didn’t want to order digitally. It was that the tools demanded they change how they behaved.
The companies were building platforms that required downloads, logins, and training. Nobody asked what retailers were doing, and what they were already doing was WhatsApp.
Built around real behaviour
nFuse, the Sofia-based startup Ivanov co-founded in 2025with fellow Coca-Cola executive Stefan Radov in 2025, has raised $2 million from Eleven Ventures and LAUNCHub to build on that insight.
Radov offers a similarly close look at why those legacy systems failed. He spent nearly a decade at Coca-Cola as Global Product Manager and later CEE Digital Manager, directly overseeing the kind of B2B platforms nFuse is now trying to displace. In other words, he watched well-funded digital tools get built, deployed, and quietly abandoned by the retailers they were supposed to serve.
With nFuse, retailers can place orders by sending a text, a voice note, or a photo of an empty shelf. There’s no app or login needed, as the system handles the rest. An AI layer interprets the message, extracts the order, and routes it into the brand’s existing fulfilment stack.
When a shop owner sends a WhatsApp message, typed in full, dictated as a voice note, or just a photo of a nearly empty shelf, nFuse’s system uses natural language processing to parse the intent. A photo of a depleted Coca-Cola bay gets interpreted as a replenishment request; a voice note in Bulgarian or Spanish gets transcribed and matched against the retailer’s order history and the brand’s SKU catalogue.
The output is a structured order that drops into whatever ERP or distribution system the FMCG brand is already running, without requiring them to rebuild their backend. The AI also learns from each interaction. Over time, it starts anticipating reorder patterns and can prompt retailers proactively rather than waiting for them to reach out
The results, the company says, are stark. Where traditional platforms see roughly 15% adoption, nFuse is achieving above 70%. Implementation takes eight weeks instead of 18 months. Revenue per outlet climbs 15–30%. The cost per order drops to under $1, up to 20 times cheaper than legacy systems.
Where rivals like Pepperi built platforms retailers are asked to adopt, nFuse is built around tools retailers already use.
What comes next?
The company is currently working with clients across the beverage, dairy, pet food, and wholesale sectors. The $2 million raise will fund expansion into more European markets, with Latin America to follow.
The longer-term bet goes beyond ordering. Every message, image, and transaction creates demand data that most FMCG brands currently don’t have in real time. nFuse plans to layer payments and micro-lending into the same conversation thread, so a retailer who places an order Monday morning could eventually settle the invoice and access a short-term credit line without ever leaving the chat window.
“As we scale, diversity in hiring across gender, background, and geography will be a deliberate priority. We’re building a team that reflects the markets we serve – multilingual, cross-cultural, and grounded in real-world FMCG experience rather than purely technical backgrounds,” notes nFuse to TFN.
The early numbers are credible, the founders have relevant scar tissue, and the core premise, meet people where they already are, is the kind of obvious-in-hindsight idea that tends to look smart later.
For now, nFuse is small, early, and operating in one of the messier corners of global commerce. That’s usually when the interesting bets get made.
