Stefan Roebel transformed from 12 years in the German military and a decade scaling Amazon operations to co-founding Europe's largest autonomous military vehicle production facility in just 3 years.
In this episode of Rockets and Radars, Stefan reveals how ARX Robotics (ARX) went from a university research project to life-saving robots supporting Ukrainian troops in under 3 years. From a tired Thursday dinner where "nothing existed" to raising €40M+ and building combat-proven systems, this is the raw story of Europe's race to build its defense future before it's too late—and how ARX could become Europe's next defencetech unicorn.
Want to get hired in ARX? https://tally.so/r/3qqAYd
Want to invest in ARX? https://tally.so/r/mZA6Ga
-----------------------------------------------
Chapters:
(00:00) Introduction
(01:57) Personal Background & Joining ARX as 4th Co-founder
(07:49) Project A Venture Studio Program
(12:40) The First Prototype
(19:37) Pivot to Modular Approach & ARX Framework
(25:45) Customer Acquisition Strategy
(31:27) Seed Round: €9 Million & NATO Innovation Fund
(39:54) Launching Mithra OS
(42:19) Series A: How we raised €31 Million
(46:24) Vision: Becoming a European Defence Prime
(50:35) Advice for Defence Tech Founders
-----------------------------------------------
Takeaways:
1) Venture studios beat accelerators for hardware companies
"A venture studio is a level two accelerator. They assign a founder associate, shared resources like hiring, legal contacts - they have all the tools a classical VC has plus operational background." This mindset was exactly what ARX needed for complex hardware development.
2) Build your first prototype to fundraise, not to sell
"Investors got super excited because they saw a physical product after a decade of SaaS presentations." Hardware founders need something tangible to overcome investor skepticism.
3) Pitch deck hell is inevitable - embrace the chaos
"We had 40 iterations of the first pitch deck.
There are 500 different opinions on pitch decks, especially for hardware. Don't fight the feedback loop - use each iteration to find investors who actually understand your market and hardware.
4) Win credibility through real competitions, not demos
"We won the visitors award at Estonian autonomy trials. That was the first time we got exposure to other robotics companies." Public trials validate your tech against established competitors.
5) Pivot from single-use to modular approach
"We found most robots are single-use. But the change of situation is so rapid you need a new solution every five minutes." Undefined markets require maximum flexibility.
6) Start with R&D contracts, scale to procurement later
"We used smaller R&D programs to finance development of 1-10 robots. The procurement is still a tanker - we're the speedboats getting money to figure out new processes." Use innovation programs as stepping stones to larger contracts.
7) Hire military experience locally for each country expansion"My military English is bad. We're replicating what we did in Germany by hiring local teams with military experience to understand customer needs and adapt our system." Defence is hyper-local - you need native military speakers.
8) Series A requires proof, not just vision"Series A is proof of concept done, show us traction, pipeline, forward-looking revenue. It's more data-driven, unromantic. They need conviction the money scales an already working machine." Move from founder-market fit to product-market fit evidence.
9) Speed of iteration determines battlefield winners"Fast iteration cycles will be the difference maker. The current speed of change is mission critical to battlefield success. Whoever iterates fastest wins."
10) Pick investors who understand B2G
]We had investors with SaaS mindset asking about recurring revenues. Find investors who've lived your customer's world, not just scaled software companies.