Today we're starting poneva. We're a small team with one conviction: shopping online is about to change more in the next few years than it has in the last twenty, and most stores have no practical way to keep up. We're here to fix that.
Why now
Large language models can finally understand a shopper the way a great salesperson does — the half-formed question, the real constraint behind it, the trade-off they haven't named yet. The technology to have that conversation exists. What's missing is the bridge between a model and a merchant's actual catalog, policies, and storefront. Building that bridge by hand is a project most teams can't afford to take on.
What we're building
poneva aims to be the best helper for e-commerce teams building integrations to LLMs. Point us at a store and we do the hard part: learn the site, ground an assistant in real products, and put a working shopping assistant on the storefront — without an integration project, a data export, or an engineering sprint.
- For shoppers: a store you can talk to, that actually knows what it sells.
- For merchants: the upside of AI on your store without the build.
Making the world agentic
We think the web is heading somewhere specific: from pages you read to agents that act on your behalf. Commerce is where that shift will be felt first and hardest — it's where intent meets a transaction. If we do our job, every store becomes a place an agent can reason about and act in. That's the world we want to help build, one merchant at a time.
It's day one. There's a great deal to prove, and we'll share the journey here — the wins, the dead ends, and what we learn along the way. Thanks for being early.