Munich-based Interloom has raised $16.5 million in venture funding to solve what founder and CEO Fabian Jakobi calls the biggest blocker to enterprise AI agent adoption: the 70% of operational decisions that have never been formally documented, Fortune reported Monday.
The round was led by DN Capital with participation from Bek Ventures (whose partner Mehmet Atici previously backed UiPath) and existing investor Air Street Capital. Interloom previously raised a $3 million seed round in March 2024. The company did not disclose its post-funding valuation.
The Problem: Smart Agents, No Context
Every enterprise AI agent deployment hits the same wall. The agent can reason, search documents, and execute workflows, but it has no idea how the specific company actually operates. The workarounds that veteran staff use, the right internal team for a particular escalation path, the fact that the documentation for a critical process is wrong: none of this is captured anywhere a model can access.
“The most important person at the bank is the person who knows whether the documentation is right or not,” Jakobi told Fortune. “They’re often the lowest paid. But they determine quality.”
Interloom’s approach is to ingest millions of operational records (support emails, service tickets, call transcripts, work orders) and construct what it calls a “context graph,” a continuously updated map of how problems actually get resolved within a given organization. Jakobi compares it to Google Maps: the graph learns optimal resolution paths from real operational data the same way navigation software learns routes from traffic patterns.
Live Deployments at Commerzbank, Volkswagen, Zurich Insurance
Interloom has already deployed with several large European enterprises. At Commerzbank, the company analyzed millions of customer support emails against internal documentation, finding that much of the documentation was either conflicting or incomplete. Interloom says it reduced the gap between documented and actual operational knowledge from roughly 50% to 5%.
At Volkswagen, the system processes customer support tickets. At Zurich Insurance, Interloom won a company-wide AI competition, beating out what Jakobi says were 2,000 other AI-native startups, for an underwriting use case. Insurance underwriting is a particularly strong test case because decisions reflect a firm’s specific risk appetite, its accumulated experience with certain brokers and products, and institutional knowledge that no general-purpose model can replicate.
“The Zurich underwriter knows how their broker chat underwriting works much better than Accenture does,” Jakobi said, positioning Interloom against the consulting firms that have traditionally dominated enterprise process work.
Timing and Competition
A wave of Boomer retirements adds urgency. Roughly 10,000 Americans turn 65 daily, taking decades of institutional knowledge with them just as companies try to deploy AI at scale. Without a way to capture that knowledge before it walks out the door, AI agents will keep hitting the same context gap.
Interloom’s next product push is an internal “Chief of Staff” layer designed to give managers real-time visibility into how their AI agents perform, including version control for agent-driven processes. The company competes with a crowded field: OpenAI, ServiceNow, and Microsoft are all building agent management and orchestration layers. Jakobi argues that Interloom’s context graph, which maps knowledge across entire complex processes rather than individual workflows, gives it a structural advantage that horizontal platforms lack.
DN Capital partner Guy Ward Thomas echoed the thesis: “An agent is only as good as the expert decisions it can rely on.”