Build the decision layer.
A small team in the West, building the intelligence that reads autonomous behaviour faster than it executes. The problems are unsolved, the instruments are imperfect, and the interesting work starts where the manuals stop.
- 01
You reason from first principles.
The reference manual runs out exactly where this work begins. We care less about what you already know than about what you do when nothing in the literature fits.
- 02
You are dangerous with little.
Cheap hardware, incomplete data, no budget for the obvious solution. The people who thrive here make modest instruments do immodest things.
- 03
You finish.
Ideas are abundant and cheap. Something that runs, in the field, on a bad day, is neither. We hire for the second one.
- Signal Processing EngineerRF · radar
- Machine Learning Engineersensor data
- Systems Engineerfield integration
- Machine Learning Engineeradversarial autonomy
- Embedded Systems Engineerreal-time
- Radar Systems Engineerhardware
- Field Integration Engineerdeployment
Nothing open at this post. Answer the three questions anyway — we keep interesting people on file.
Signal processing & sensor fusion · Machine learning on sparse, adversarial data · Embedded & real-time systems · Systems engineering · field integration. No matching title? Answer the three questions anyway — we hire people, not job descriptions.
No CV. No cover letter. Answer three questions. There are no correct answers — only revealing ones. We read how you reason when the signal fights back, and we get back to you within 72 hours.
We are not looking for the textbook answer. We are looking for how you think when the signal refuses to cooperate.
Invent the tell. Argue for it.
Creativity over correctness. Surprise us.
Opens your mail client — nothing is stored on this site. Prefer your own mail? Write to contact@bellum.systems and answer the three above.