Remote
Senior Data Scientist
About this role
Swarmer develops software that makes drones autonomous and allows them to operate together, in large, coordinated teams — no pilots needed. Our technology has been battle-tested in Ukraine— the world’s most intense proving ground for drone warfare. In March 2026, we became the first Ukrainian defense startup to go public on NASDAQ, following a $15M Series A — the largest investment in a Ukrainian defense tech company since the start of the war.
Working at Swarmer means operating at the intersection of engineering rigor and frontline reality. The problems and environments are complex, and the stakes are very real. We built this software to enable democratic nations to defend themselves. If you are motivated by building resilient systems that matter and by seeing the direct impact of your work, you’ll find purpose here. Who we are looking for: We are looking for a hands-on Data Scientist who can own and contribute to existing ML/CV pipelines in a robotics autonomy stack, and who can also own and grow the data side of the house — from collection and labeling through lakes/warehouses to training and evaluation.
This is an IC-first role with real ownership: you write code, ship models and data products, and improve the loops that turn flight and simulation data into better autonomy. We are not looking for a pure researcher, a detached platform architect, or someone who only does notebooks. We need someone who can work end-to-end across models, data, and robotics constraints, with strong judgment and delivery discipline. What you’ll do: - Own and contribute to existing ML/CV flows in our robotics autonomy and perception stack (LMT, ATR, VISNAV and related systems): training, evaluation, iteration, and integration with onboard/edge pipelines.
- Design, build, and operate data flows: ingestion from flights and sims, storage (data lakes / warehouses), labeling workflows, dataset versioning, and reproducible training/eval pipelines. - Work extensively with simulated data: generate, validate, mix with real flight data, measure sim-to-real gaps, and improve dataset quality for perception and autonomy models. - Improve model quality under field constraints: noisy sensors, limited compute, edge deployment (e.g.