AI Research Engineer - ML & Signal Processing

At Helsing you will work at the intersection of machine learning and signal processing on RF data, helping build the AI capabilities inside our electronic warfare systems: models that detect, classify, and reason about RF signals in congested, contested spectrum. The data is messy, irregular, and adversarial. Output runs on real platforms, under hard latency and compute budgets. What you build has to survive contact with the deployment target. You will own problems end-to-end: read the literature, prototype on real recorded data, push to the edge, watch it run. You will work in a small research team alongside RF and hardware engineers and the operators using the system. Current focus areas include self-supervised representation learning on signal data, deinterleaving and emitter identification, mode recognition, and anomaly detection on embedded hardware. You should apply if you hold a PhD (or have an equivalent research track record) in ML, signal processing, physics, electrical engineering, or a related field, with publications at NeurIPS, ICLR, ICML, ICASSP, or similar venues. Have shipped non-trivial ML into a real product - not just notebooks. You know what breaks between research and deployment. Are comfortable on irregular, noisy sequence data: temporal models, self-supervised or contrastive methods, calibration under distribution shift. Write clean Python, and are willing to pick up Rust where the deployment target demands it. Read papers and reimplement them. You can tell which results will hold up and which won't.