n8n

Jan Oberhauser did not come from enterprise software. He came from visual effects studios — Digital Domain, Pixomondo, Neon — where he worked as a pipeline TD and compositor, the person responsible for making sure that the dozens of software tools, render farms, and asset management systems inside a VFX production house actually talk to each other. That problem — getting heterogeneous systems to exchange data reliably, without writing bespoke glue code for every new pairing — became his obsession. He tried to solve it with link.fish, a web scraping and data extraction startup. Then in 2019, working alone in Berlin, he built the more fundamental tool he actually needed: a node-based automation platform where any application could connect to any other application with no limits on complexity. He put it on GitHub. The community found it. And a slow, compounding flywheel started turning.

For six years, n8n grew steadily — 3,000 enterprise customers, 200,000+ active users, $74M raised across seed, Series A, and Series B. Then the AI wave arrived. In 2022, Oberhauser made the call to rebuild n8n around LLMs and AI agents as first-class citizens — not as a feature bolted onto the side, but as the primary way workflows would be designed and executed. The result was four times the revenue growth in eight months compared to the prior six years combined. By August 2025, ARR had reached $40M. In October 2025, Accel led a $180M Series C at a $2.5B valuation, with NVIDIA's NVentures and Deutsche Telekom's T.Capital joining as strategic investors. The United Nations runs mission-critical workflows on n8n. So does Microsoft, Vodafone, Wayfair, Meta, Zendesk, and Delivery Hero. There are 173,000+ stars on GitHub. The platform has 500+ integrations and ships MCP client and server nodes — meaning n8n workflows can now both consume and expose tools to AI agents from Claude, Cursor, and any other MCP-compatible system. Oberhauser's framing for where this goes: "We're building for a future where AI-driven automation becomes as essential as Excel once was."