INXM emerges from stealth with €5.7M pre-seed to build an AI process execution engine
What's the deal? INXMDealroom has a profile for this one. Try Dealroom →, a Berlin-based startup building an AI process execution engine for enterprises, has emerged from stealth with €5.7 million in pre-seed funding. Cherry Ventures and RedstoneDealroom has a profile for this one. Try Dealroom → led the round, with participation from Angel InvestDealroom has a profile for this one. Try Dealroom →, Linden CapitalDealroom has a profile for this one. Try Dealroom →, and other angels.
Founded in 2025 by Alex Oelling, Matthias Kainer, Jesper Bylund, and Kamil Klüber, INXM is targeting a persistent pain point: enterprise AI projects that take years, require armies of engineers, and break more than they fix. The company's answer is what it calls "compiled AI" — using large language models to design processes, then running them as deterministic, auditable code.
"Knowledge workers still copy-paste between ERP, PLM, Excel, email, and approval workflows to close a month. We have set out to build AI that finishes the work for you," said Oelling, the company's chief executive officer.
Why now? Enterprises — especially Germany's Mittelstand — are under pressure to adopt AI but keep hitting the same wall: LLM-powered tools are flexible but unreliable in production. INXM's core bet is that separating the intelligence layer (process design) from the execution layer (deterministic code) solves this. The company claims its Orchestrator can automate processes within months, not years, without ripping out existing systems.
"At its core, compiled AI means you use LLMs to generate deterministic, enterprise-ready code. You then run the code to achieve your outcome," said Kainer, the company's chief technology officer.
What could go wrong? The "compiled AI" framing is novel, but INXM enters a crowded field of enterprise process automation tools — from established players like UiPath and Celonis to a wave of AI-native startups. Convincing risk-averse European enterprises to trust a pre-seed startup with core operational workflows is a steep climb, even with a fast deployment promise.
The company will also need to prove that its deterministic execution approach works across the messy reality of heterogeneous enterprise tech stacks.
The signal: Cherry Ventures and Redstone both backing a pre-seed round of this size underscores growing investor appetite for enterprise AI infrastructure that prioritises deterministic reliability over raw model capability. With INXM already tagged at "early growth" stage on Dealroom despite just exiting stealth, the bet appears to be on rapid enterprise traction — particularly among Germany's Mittelstand, where process automation demand is high but tolerance for unpredictable AI outputs is low.
Read more: EU-Startups
