The Munich startup is building what it calls a ‘context graph’, a continuously updated map of how operational decisions actually get made inside an enterprise, drawn from millions of real cases rather than documentation that may never have been written.
There is a particular friction point in every enterprise AI deployment, and anyone who has tried to roll out an AI agent in a large organisation tends to run into it early. The agent might be technically capable. It can read documentation, follow instructions, and execute steps.
What it cannot do is replicate the judgment of the person who has been doing the job for fifteen years and knows, from experience, exactly why the standard playbook does not work on Tuesdays in the logistics department. That knowledge has never been written down, because nobody needed to write it down, until now.
This is the problem Interloom is building towards. The Munich-based startup, which operates across Munich, Berlin, and London, announced on 19 March that it has raised $16.5 million in a seed round led by DN Capital, with participation from Bek Ventures and existing backer Air Street Capital.
The round represents a significant step up from the company’s initial $3 million seed, which Air Street led in March 2024 when the company emerged from stealth.
Interloom’s core product is what it calls a Context Graph: a continuously evolving model of how operational decisions actually get made inside a given organisation, constructed by ingesting millions of real cases, support emails, service tickets, call transcripts, work orders, and extracting the patterns of how expert workers resolve problems.
Founder and CEO Fabian Jakobi describes the challenge in terms of tacit knowledge, the concept coined by British-Hungarian philosopher Michael Polanyi whose central observation was that most expertise cannot be fully articulated by the expert who holds it. Jakobi estimates that around 70% of operational decisions are never formally documented.
The analogy Jakobi uses is Google Maps: just as the navigation tool learns optimal routes from real-time traffic, Interloom builds a map of the paths operational experts actually take to solve problems, then uses that map to guide AI agents and new employees facing similar situations. The system updates continuously, so that every resolved case adds to the institutional memory rather than disappearing when the person who handled it leaves or retires.
That retirement risk is part of the pitch. The press release cites the figure of 10,000 baby boomers leaving the US workforce daily, a demographic statistic widely documented by Pew Research.
The argument is that enterprises face a compounding problem: institutional knowledge built up over decades is being lost at precisely the moment AI is expected to step in and automate complex operational work. Without capturing that knowledge first, the AI has nothing useful to draw on.
Interloom’s early customer base includes Zurich Insurance, JLL, and logistics group Fiege, as well as Commerzbank and Volkswagen, the latter two confirmed independently by Fortune in its exclusive on the funding. At Commerzbank, Interloom analysed millions of customer support emails against internal documentation and reportedly reduced the gap between what was documented and how work actually happened from around 50% to 5%.
At Zurich Insurance, the company won an internal AI competition against what Jakobi described to Fortune as 2,000 competing AI-native startups for an underwriting use case.
The investor lineup carries its own thesis-confirming logic. Guy Ward Thomas, the DN Capital partner leading the investment, was previously the first institutional backer of Cognigy, the German enterprise conversational AI platform, which DN Capital backed from its Series A in 2019 and which was acquired by NICE in August 2025 for $955 million — described at the time as Europe’s largest AI exit.
Ward Thomas has noted that the fundamental lesson from that investment was how critical organisation-specific context is to making AI agents work in practice. Mehmet Atici, who leads the Bek Ventures participation, was an early backer of UiPath, the robotic process automation pioneer that listed in New York in 2021. His argument is that the current wave of AI agent adoption represents the next major inflection in enterprise automation after RPA.
