Last year, my team and I at Studio were building an AI teacher — personalized curriculums for learning anything — and we wanted to add relevant product recommendations.

To build that experience, we needed real-time product data across the web. But search APIs returned articles or stale PDPs, and indexing and understanding every product online clearly wasn’t a feature — it was an entire company.

George and I decided to build it. Today, I’m excited to share that Channel3 has raised a $6M seed round led by Matrix, joined by Y Combinator and Ludlow Ventures.

Channel3 is building a universal, AI‑ready product graph – a connected database of every product on the internet. We're giving developers a simple API to find any product online – and earn commission on sales – and helping brands win incremental revenue from this new channel.

Humans shop on Google; AIs shop on Channel3

Shoppers ask AI assistants to “find a durable commuter backpack” or “style my bedroom.” Those experiences only work if the AI can see the world of products clearly: what items exist, who sells them, which variants are available, how they’re priced and stocked across merchants, and where to send the shopper to buy. Channel3 lets AI agents find any product online, and earn commission when they drive sales.

Developers get:

• An API to find any product on the internet.
• Real‑time data that stays fresh as catalogs change.
• Built‑in monetization via affiliate rails, so they earn commission on sales.

Brands get:

• Distribution into AI shopping surfaces with no integration work.
• Visibility into what people search for and purchase.
• Control over how they show up on AI shopping platforms.

Why now

Product data is messy, incomplete, and scattered across the internet. Multimodal models are finally smart enough to let us make sense of it – and inexpensive enough to do so at scale.

At the same time, AI‑native shopping is emerging as a third distribution channel alongside stores and e‑commerce. That creates a shared need for a neutral product graph that any AI app or agent can use.

Where we are today

Channel3’s catalog already spans more than 50 million products across tens of thousands of brands, and it’s growing quickly. Our pipeline uses multimodal AI to read PDPs, match duplicates across merchants, link variants like color or size, and extract rich attributes – so agents can reason about products the way a person would. Hundreds of developers are building on our API to power stylists, interior design tools, gifting agents, and shopping assistants. Most products in our catalog are automatically monetizable via partnerships with the major affiliate networks.

What’s next

We’re laser-focused on three goals:

1. Build the authoritative source of product data on the web
2. Let anyone build agentic shopping experiences
3. Help brands win sales from this new channel

Huge thanks to our investors for believing in this vision – and to the team bringing it to life: my co‑founder, George Lawrence, and our founding engineers, Evan Fenster and Ignacio Valdez.