Marble Raises $3.75M for Observational AI Coworkers

Marble Cognition, an emerging player in AI-driven business process automation, has successfully closed a $3.75 million pre-seed funding round. The investment was co-led by prominent venture capital firms Heartcore and Seedcamp, signaling strong confidence in the startup's innovative approach to augmenting workforce capabilities. The round also saw participation from a distinguished group of angel investors, including a board member from Datadog, the founder of Personio, and key executives from industry leaders such as McKinsey, Mistral AI, and Lovable.

Founded in late 2025 by Felix Rösner and Christophe Schmit, Marble is tackling a critical bottleneck in enterprise AI adoption: the lack of structured understanding of internal workflows. Many essential business processes remain undocumented, with knowledge dispersed across disparate systems and employee tacit understanding. This opacity hinders the effective identification and automation of tasks, despite the theoretical potential for AI to automate a significant portion of work hours – estimated by McKinsey to be around 57% in the US. Marble's research indicates that approximately 90% of core business processes within companies are not formally mapped, presenting a substantial challenge for AI integration.

Marble's core innovation lies in its ability to observe and learn from human employees as they execute their daily tasks. By analyzing user interactions across business applications, including screen activity, clicks, and keystrokes, the platform constructs detailed, interpretable process maps. This observational learning approach allows Marble to develop AI agents, dubbed 'Marble Coworkers,' capable of replicating these workflows. Crucially, the system prioritizes data privacy, anonymizing information and analyzing it in aggregate, not at an individual level. Employee monitoring can be paused, and the application is designed to avoid tracking private applications or activity outside of work hours.

The 'Marble Coworkers' are designed to seamlessly integrate into existing organizational toolchains, executing business processes autonomously. This allows human employees to focus on tasks requiring higher-level judgment, strategic decision-making, or final approvals, thereby enhancing overall productivity and efficiency. "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. Our Coworkers watch how the work happens before they take it over. They do it the way your best people do," stated Felix Rösner, co-founder of Marble.

The company has already demonstrated the efficacy of its technology through deployments at several organizations, including EGYM, Mutares, and Enam, alongside other technology firms and private equity-backed entities. This early traction validates Marble's solution in real-world business environments. The newly acquired capital will be instrumental in scaling Marble's inference capabilities, bolstering its engineering team based in Munich, and facilitating its strategic expansion into the United States market.

This funding round underscores the growing investor appetite for AI solutions that address practical enterprise challenges. As businesses increasingly seek to leverage AI for operational efficiency, startups like Marble, which offer a tangible path to automation through intelligent observation and learning, are well-positioned to capture significant market share. The ability to translate undocumented human processes into machine-executable workflows represents a key advancement in the field of business process automation.