Ex-Isar Aerospace and Volocopter leaders raise €5.7M to fix the enterprise AI execution problem — TFN

• INXM has raised €5.7M in pre-seed funding led by Cherry Ventures, with participation from Redstone, Angel Make investments, and Linden Capital, to construct what it calls the primary AI Course of Execution Engine for enterprise operations.
• The Berlin-based startup, based in 2025 by former CDOs of Isar Aerospace and Volocopter alongside veterans of n8n and Thoughtworks, is focusing on industrial producers struggling to show AI-generated workflows into dependable, auditable enterprise outcomes.
• European enterprise AI infrastructure is attracting rising investor consideration as organisations transfer from AI pilots to manufacturing deployments and demand repeatability over experimentation.

Enterprises have spent billions on AI. Most of that spending has produced dashboards, copilots, and pilot initiatives, not dependable manufacturing workflows. The hole between AI producing a plan and an enterprise constantly executing one is the place INXM was constructed to function.

The Berlin-based startup has emerged from stealth with €5.7M in pre-seed funding led by Cherry Ventures, alongside Redstone, Angel Invest, and Linden Capital. The capital will assist the business rollout of INXM’s flagship product, Orchestrator, and deepen improvement of its core know-how platform.

The founders

INXM was based in 2025 by Alex Oelling (CEO), Matthias Kainer (CTO), Jesper Bylund, and Kamil Klüber. Oelling beforehand served as Chief Digital Officer at each Isar Aerospace and Volocopter, the place he constructed the organisational and digital infrastructure behind a few of Europe’s most demanding aerospace programmes. Kainer labored alongside Oelling at each corporations, constructing mission management programs, launch orchestration, and cloud-native replacements for legacy operational platforms. The crew’s backgrounds additionally span n8n and Thoughtworks.

“We based INXM as a result of we’ve seen first-hand how enterprise AI initiatives fail: years of implementation, armies of engineers, and AI programs that break greater than they repair. We’ve got down to construct AI that finishes the give you the results you want — the system that turns AI from a productiveness software into the operational spine of European trade,” mentioned Oelling.

What INXM truly does

On the centre of the corporate’s providing is Orchestrator, constructed round what INXM calls “Compiled AI.” Moderately than permitting an LLM to make real-time selections all through an operational course of, the agent mannequin attracting a lot of at present’s AI funding, INXM, makes use of AI to design and optimise execution plans beforehand. These plans are then executed deterministically, guaranteeing outcomes stay repeatable, auditable, and compliant with enterprise governance necessities.

“At its core, Compiled AI means you utilize LLMs to generate deterministic, enterprise-ready code. You then run the code to realize your consequence. This offers you the flexibleness of pure language from AI fashions, however the testability of deterministic code,” mentioned Kainer.

The platform integrates with current ERP, MES, PLM, and high quality administration programs slightly than changing them, focusing on producers, aerospace corporations, and logistics operators who can not afford unpredictable outcomes in manufacturing environments.

The traders

“Enterprise AI is caught in a paradox: the extra bold the deployment, the much less predictable the result. INXM has reframed the issue totally. It’s not about making AI smarter, however making it executable. Compiled AI is a brand new architectural paradigm, and INXM is the crew to outline it,” mentioned Filip Dames, Founding Accomplice at Cherry Ventures.

“Founders who’ve introduced rocket engines and air taxis to manufacturing readiness develop a special mindset than most software program groups. They perceive that AI workflow challenges virtually all the time hint again not solely to the fashions themselves, however to brittle integration with day-to-day operations. INXM is constructed by individuals who’ve lived that.” says Michael Brehm, Founding Accomplice, Redstone.

“Enterprises don’t want extra disconnected AI instruments. They want AI that may execute ruled, repeatable processes throughout the programs they already use. INXM is constructing the orchestration layer that makes this doable: connecting enterprise platforms, people, and AI to run operational workflows with reliability, traceability, and management at scale.” says Jens Lapinski, Founding Accomplice, Angel Make investments.

The aggressive panorama

INXM enters a market the place a number of well-capitalised gamers are constructing adjoining infrastructure. n8n, the Berlin-based workflow automation platform, raised $180M in a Series C in October 2025 at a $2.5B valuation, and has since acquired a strategic funding from SAP that pushed its valuation to $5.2B, making it one in all Germany’s most dear AI corporations. n8n focuses on connecting purposes by means of visible workflow automation; INXM targets deterministic execution with enterprise-grade governance, placing the 2 corporations in adjoining slightly than equivalent territory.

Parloa, the Berlin-based AI agent platform for enterprise customer support, raised $350M in a Series D in January 2026 at a $3B valuation, tripling in worth in underneath eight months. Parloa operates primarily in touch centre automation slightly than industrial operations. LangChain raised $125M in a Sequence B led by IVP at a $1.25B valuation in October 2025, offering foundational developer infrastructure for AI purposes and brokers, however targeted on builders constructing new programs, whereas INXM targets operational groups managing current ones.

What differentiates INXM is its emphasis on industrial environments and compliance-first, deterministic execution, areas the place enterprises stay reluctant handy management to autonomous AI programs and the place the implications of unpredictable outcomes are measured in operational threat, not inconvenience.

What’s subsequent?

Enterprise AI’s first wave was about producing content material. Its second is about deploying brokers. The query INXM is betting on is whether or not the third wave belongs to the execution layer, the infrastructure that makes AI truly end the work as soon as it begins.