Octonomy closes $20M seed, commercialise agentic AI for industry.

octonomy has closed a $20 million seed round to accelerate commercialization of its agentic AI platform that digests and acts on complex technical data. The German start-up says its software is built to interpret schematics, service manuals, ERP traces and live maintenance logs — the kind of unstructured inputs that derail many conventional AI projects.

The round was led by Macquarie Capital, with participation from Capnamic, NRW.Bank, and the TechVision Fund. The new capital lifts octonomy’s total financing to about $25 million since the business launched in early 2024 and will bankroll a rapid expansion across Europe and North America.

Founded by Sushel Bijganath (CEO) and Oliver Trabert (CPTO), together with Thorsten Grote, Markus Hanslik and Thomas Bollig, octonomy has focused from day one on creating autonomous agents that go beyond conversational assistants. The company highlights verified >95% response accuracy on technical workflows in pilot deployments — a result it contrasts with typical baseline accuracy figures seen in general-purpose chat AI.

octonomy’s proposition is operationalisation: its agents do more than answer questions — they orchestrate task-specific sub-agents to read diagrams, reconcile parts lists with ERP records and trigger procedures across enterprise tooling. That approach targets sectors where documentation volume and technical complexity block automation, notably heavy equipment, manufacturing and field services.

Deployment speed is a central selling point. octonomy reports typical integrations complete in under 20 days and without wholesale data migration, enabling teams to preserve system boundaries while unlocking automation. The start-up now employs close to 70 people, is headquartered in Cologne and has opened a US hub in Denver to drive sales and implementation in North America.

Market context favours specialised automation: industry research and consulting firms estimate that a large share of enterprise AI pilots never reach production when faced with multi-source technical data. As global firms push to translate AI research into measurable operations gains — McKinsey’s macro estimates on AI adoption remain a frequent touchstone for investors — demand for agents that can reliably execute complex tasks is rising.