DSP Engineer (Radar)
Overview
You own the signal chain inside our radar: raw ADC samples in, clean detections and tracks out, running in real time on a small TI radar chip.
Main Responsibilities
– Build the FMCW processing chain: range FFT, Doppler FFT, windowing, and angle estimation across the MIMO virtual array.
– Detect targets in clutter: CFAR detection, clutter and ground rejection, and suppression of rotor and motor noise.
– Hold up under jamming: detect interference and EW, and keep tracking through it.
– Track: Kalman filters, multi-target tracking and track management, from detections to stable track histories.
– Make it fit: map the chain onto the C66x DSP, the hardware accelerator and DMA, within strict memory, latency and power limits. Fixed-point where needed.
– Close the loop with the field: take raw data from flight tests, find what failed, fix the algorithm, and ship new firmware within days.
– Hand off clean data: deliver point clouds and tracks to the firmware and to our edgeML classifier.
– Shape the next radar: set processing needs for the next-generation (cascaded multichip) design together with the RF engineer.
Mandatory Requirements
Must have
– 3+ years of radar or similar real-time signal processing (FMCW radar, sonar, comms, EW).
– Strong DSP fundamentals: FFT, filtering, detection theory, estimation.
– Algorithms you deployed on real embedded hardware (DSP, FPGA, MCU or SoC), not only simulated.
– Strong C or C++ for embedded targets. MATLAB or Python for analysis and prototyping.
– You work from real field data, not ideal simulations.
Nice to have
– TI mmWave experience (AWR or IWR family, mmWave SDK, C66x DSP, hardware accelerator).
– Tracking and sensor fusion (Kalman, IMM, data association).
– Fixed-point optimization, DMA and memory-constrained design. Drone, UAV or defence background.
We offer
– Location: on-site in our Kyiv lab or in Vilnius.
– We go from bench to field tests in days.
– Package: top-of-market salary plus founding equity.
– Relocation support, and housing for Kyiv.
– Team: you work daily with our RF, hardware, firmware and ML engineers.
– You report to the CEO.
– Funding: we closed our pre-seed in November 2026 and work directly with Texas Instruments’ radar team.