Munich-based Interloom, which aims to capture tacit knowledge for AI agents from businesses' operational records, raised a $16.5M seed led by DN Capital
Michael Polyani, the British-Hungarian philosopher, economist, and scientist, is perhaps best known today for coining the term “tacit knowledge.”
Context & Ripple Effects
Interloom’s seed round sits within a German enterprise-AI cluster focused on operational work: Munich’s Tacto applied AI to supply-chain optimization, while Cologne-based Octonomy has targeted complex business workflows. Interloom is aimed one layer earlier, turning the implicit know-how embedded in operational records into material AI agents can use.
The broader agent market is also moving beyond scripted interaction, as NeoCognition’s self-learning agent approach illustrates. Interloom’s emphasis is not a new foundation model, but the business-specific knowledge needed to make agents useful in real operating environments.
First-order effects
- Interloom gains $16.5 million in seed capital, led by DN Capital, to build its approach to extracting tacit knowledge from companies’ operational records for AI agents.
- Prospective customers get another route to make agents reflect internal processes and accumulated operating experience, rather than relying only on general-purpose model behavior.
Second-order effects
- Enterprise-AI vendors focused on workflows will face pressure to show how their agents acquire, structure, and retain customer-specific operational knowledge—not just automate a single task.
- Implementation partners and internal AI teams may place greater value on record integration and knowledge capture, since those inputs determine whether an agent can operate in a company’s particular context.
Third-order effects
- If this model proves repeatable, proprietary operational knowledge could become a more durable competitive asset in enterprise AI than access to the same underlying general models.
- The market may increasingly reward AI-native systems that combine model capability with ongoing knowledge capture, strengthening the case for an enterprise agent deployment model over standalone conversational interfaces.
The trend: Enterprise AI is shifting from generic assistants toward agents differentiated by their ability to learn from and operate on company-specific knowledge.