Reduck secures $3M to make web automation faster for AI agents

French-American AI startup Reduck has attracted $3 million in pre-seed funding to enhance web automation for AI agents. The investment, led by Daphni, was joined by Drysdale Ventures and Galion.exe. Reduck's technology aims to enable AI agents to interact with websites and automate tasks without traditional APIs. Founded by CEO Daniel Huynh, CTO Louis Abraham, and COO Raphaël Millet, Reduck seeks to tackle the issue of limited direct access to websites and business applications, which often lack APIs or Model Context Protocol (MCP) interfaces.

As a result, AI agents typically rely on slow, costly, and unreliable browser-based interactions to perform complex tasks. To address this, Reduck developed Reduck MCP, a platform allowing AI agents to discover, execute, and create browser automation scripts, making agents three times faster and cheaper for tasks like downloading invoices and market research.

The company's library currently includes over 1,200 ready-to-run automation scripts for more than 250 websites. Reduck is preparing a developer offering to enable teams to build custom AI agent systems connected to non-API applications, including specialized portals in insurance and healthcare. The company also explores applications for automating knowledge work, starting with generative engine optimization (GEO).

The funding supports Reduck's team expansion in AI engineering, full-stack development, and infrastructure, with plans to further improve automation technology, expand the script library, and introduce a new developer offering.