Building the product data infrastructure for agentic commerce.

Co-founders Hamish Gunasekara and Dylan Farrell have spent their careers where commerce meets machine learning. Hamish, CEO, started out as an early data scientist at Afterpay; after Block acquired it, he worked across Square and Cash App, leading their merchant integrations. Dylan, CTO, studied pure mathematics at Harvard, then spent years as a staff machine-learning engineer at the fintech Curinos, where he built its product recommendation engine.

That work convinced them of something bigger. AI is changing how people use the internet: more and more, what used to start with a search or a website now starts with a question to ChatGPT. Shopping is no exception. But the systems that hold the world’s product information were built for people browsing pages, not for AI trying to make sense of what’s actually for sale.

So they started Catalog. The premise is simple: an AI system is only as good as the data underneath it, and most product data was never written for machines to read. It’s scattered across channels, locked inside marketing copy, and missing the details a model needs to reason about a product at all. Catalog takes a brand’s raw data and turns it into clean, structured product information that AI can actually use, then keeps it current as the catalog changes.

The bigger goal is for Catalog to become the operating system for agentic commerce: the rails AI uses to discover and reason about products across the open web. As shopping moves into AI, that’s the layer it will run on.