Reduck Raises $3M Pre-Seed Led by Daphni to Give AI Agents Ready-Made Scripts Instead of Slow, Step-by-Step Clicking
Reduck Daniel Huynh (CEO), Louis Abraham (CTO), and Raphaël Millet (COO)
Image credit: Reduck
Reduck, a French‑American AI startup, has raised $3 million in a pre‑seed round to make web automation faster and cheaper for AI agents. Daphni led the round, with participation from Drysdale Ventures and Galion.exe. The company develops technology that lets agents work with websites and business applications that do not offer an API.
Reduck was founded in 2026 by Daniel Huynh, chief executive, Louis Abraham, chief technology officer, and Raphaël Millet, chief operating officer. It operates in France and the United States.
AI agents are increasingly built to automate digital tasks, but many websites and internal business tools do not provide APIs or Model Context Protocol interfaces, the connections that let an agent call a service directly. Without them, agents usually fall back on browser‑based interaction, clicking and typing through a page as a person would. That approach can be slow, costly and unreliable when the task is complex.
Reduck has built Reduck MCP, a platform through which agents can discover, execute and create browser automation scripts. Rather than navigating a website one step at a time, an agent uses a predefined script to complete the task, which reduces the time and computing resources involved.
According to the company, this approach can make agents three times faster and cheaper on tasks such as downloading invoices from multiple vendors and conducting market research on LinkedIn. The figure is the company's own and has not been independently verified.
Reduck's library currently includes more than 1,200 ready‑to‑run automation scripts covering over 250 websites. To grow and maintain that library, the company is building a system of AI agents that can write and update scripts on their own, with a focus on reliability and scalability.
Reduck is preparing a developer offering that will let teams build custom AI agent systems connected to applications without APIs. The examples it gives include specialised portals in insurance and healthcare as well as internal customer relationship management systems.
The company is also exploring uses in knowledge work, starting with generative engine optimisation, a practice aimed at improving how brands appear in answers produced by AI systems.
The new money has supported expansion of Reduck's team across AI engineering, full‑stack development and infrastructure. The company plans to improve its automation technology further, add to its script library and introduce the new offering for developers. It did not disclose a valuation.
Huynh, Abraham and Millet lead the company. Reduck's website is reduck.ai. Daphni is the lead investor, and Drysdale Ventures and Galion.exe joined the round.
The company describes the gap it addresses as one of the obstacles to wider use of AI agents: agents often lack direct access to the sites and applications where work happens. Reduck's answer is to give them reusable scripts rather than asking them to work out each page from scratch.
Browser automation is one route by which agents reach software that exposes no other way in. An agent can look up a script that already covers a given site and task instead of working through the pages again. The company has not published pricing or customer names.
Topics
Stay informed
Startup news in your inbox
Get important funding rounds, founder stories, and startup updates.
No spam - only important startup updates.
Comments
Text only. No links or images.
No comments yet.
