Want to connect with Catalog?
Join organizations building the agentic web. Get introductions, share updates, and shape the future of .agent.
Is this your company?
Claim this profile to update your info, add products, and connect with the community.
Catalog is a core infrastructure provider for the agentic commerce sector. While much of the AI ecosystem focuses on the reasoning capabilities of agents, Catalog provides the reliable data foundation those agents need to interact with the retail world. Their platform is specifically designed to solve the "messy data" problem that prevents LLMs and autonomous agents from accurately identifying, comparing, and purchasing physical goods.
They are active participants in the development of machine-to-machine commerce protocols like ACP and UCP. This makes them highly relevant to developers building shopping agents or retailers who want to ensure their products are discoverable within the growing number of AI-driven interfaces. By providing deterministic data objects and real-time synchronization, they enable agents to move from simple recommendation tools to reliable transaction executors.
The way people buy things online is moving away from the traditional model of browsing category pages and comparing tabs in a browser. Instead, consumers are increasingly using large language models to describe their intent in plain language. "What are the best headphones for travel?" is a query that requires an AI to not only understand the user's context but also to filter, compare, and recommend products accurately. Catalog is built on the premise that current e-commerce infrastructure is unprepared for this change. Most product data was designed for human eyes, existing as messy text, inconsistent variant tables, and unreliable availability signals that break when an agent tries to parse them.
Founded in San Francisco in 2025 by Hamish Gunasekara and Dylan Farrell, Catalog builds the data infrastructure that makes products readable for AI systems. Gunasekara, formerly a data scientist at Square and Cash App, and Farrell, a staff ML engineer with a background in recommendation engines, have designed a platform that bypasses traditional integration hurdles. The system extracts product data directly from merchant websites and commerce platforms without requiring a complete schema redesign.
Once the data is ingested, Catalog normalizes it. This process involves resolving inconsistencies in sizes, colors, and materials to produce what the company calls deterministic product objects. Unlike a standard search result, which might provide a snippet of HTML, these objects are structured specifically for machine reasoning. This ensures that when an AI agent makes a recommendation, it is doing so based on accurate, verified attributes rather than hallucinated or poorly parsed web text.
Catalog is active in the push for standardized protocols like the Agentic Commerce Protocol (ACP) and Universal Commerce Protocol (UCP). By distributing product data through these channels, the company ensures that merchants can show up in AI interfaces with the same level of visibility they once sought through SEO. This is a critical move as AI becomes the primary storefront, filtering options and potentially completing transactions on behalf of users.
The platform also handles continuous synchronization. In a world where an agent might be authorized to make a purchase for a user, pricing and stock accuracy are no longer optional—they are technical requirements. Catalog monitors changes in real time so that the AI systems accessing their data are always operating on the latest information. This reduces the friction that occurs when an agent recommends a product that is out of stock or incorrectly priced.
With a $3 million pre-seed round led by Acrew Capital, Catalog is scaling its engineering team to deepen its connection with major AI platforms. They are building toward a vision where the agent is the interface for the real economy. By providing the translation layer between raw merchant catalogs and structured product intelligence, they are securing a position as a necessary intermediary in the emerging agentic commerce sector. Their focus on high-fidelity data objects distinguishes them from legacy scraping services that prioritize volume over machine-readable precision.
A platform that turns merchant catalogs into structured, machine-readable data for AI discovery and transactions.
Curated Agent Skills for Microsoft & Azure – giving AI coding assistants structured, real-time expertise from Microsoft Learn docs.
Official Microsoft Learn MCP Server and CLI tool – powering LLMs and AI agents with real-time, trusted Microsoft docs & code samples.
Catalog is hiring
You've explored Catalog.
Join organizations building the agentic web.