Marble raises $3.75M to build AI coworkers that learn by watching
Munich-based Marble Cognition has secured $3.75 million in pre-seed funding to build AI agents that observe how employees work and then automate those processes.
Marble Cognition, a startup building AI agents that learn business processes by observing employees, has raised $3.75 million in a pre-seed round co-led by Heartcore and Seedcamp. Angel investors also took part, including a board member of Datadog, a founder of Personio, and executives from McKinsey, Mistral AI and Lovable.
Founded in late 2025 by Felix Rösner and Christophe Schmit, Marble targets a common obstacle to workplace automation: companies often do not know how their work is actually done. Key knowledge is scattered across emails, spreadsheets and the heads of employees, so it is hard to tell which tasks can be automated and how an AI system should carry them out.
The scale of the gap is notable. McKinsey estimates that existing technology could theoretically automate about 57 per cent of working hours in the US. Marble says that among its customers, around 90 per cent of core business processes are not formally documented.
How Marble Coworkers work
The platform analyses how staff perform daily tasks and turns them into structured process maps that AI can interpret. From these, it builds agents called Marble Coworkers, which run processes using the tools a company already has, while leaving judgement calls and approvals to people.
Observation happens through a desktop application that tracks workflows across business apps, including screen activity, clicks and keystrokes. According to Marble, data is anonymised before employees or AI models can access it and is analysed in aggregate, not per individual. The app does not monitor private applications, private browser tabs or activity outside working hours, and employees can pause it at any time.
"Every company we talk to wants AI to take over real work. But nobody can tell the AI how that work is actually done. So we solve that first," said Rösner. "Our Coworkers watch how the work happens before they take it over."
What comes next
Marble says its technology is already deployed at organisations including EGYM, Mutares and Enam, as well as other technology companies and private equity-backed businesses. The new funding will go towards scaling its inference capabilities, growing the engineering team in Munich and launching in the US market.
Apps worth a look
SUMO: the number puzzleTap the right tiles, add them up, hit the target. Easy to learn, hard to master. Free, for all ages.
Dialo: learn languages by speakingTalk to an AI conversation partner; every sentence comes back corrected and explained in your language. 9 languages, A1 to C2.
ediviewer: read EDI files like documentsOpen EDIFACT and X12 messages, see every segment explained, convert to JSON/XML, export PDF and Excel. Runs locally, no upload.
More from this day
See all →
Gallium Defense Raises €7.1M for Low-Cost Drone Detection Radar
SuperMe raises $20 million to connect AI profiles of experts
BMW to Launch Level 2 Urban Driving Assist in Germany in 2027
