About us: We are an international defence startup, rapidly expanding to support the Ukrainian national needs. As a startup, we operate in a highly flexible environment where agility, adaptability, and a hands-on approach are critical for success..
Key Responsibilities
- Research and develop computer vision tailored to defense and aerospace applications, including object detection, classification, and tracking.
- Design and optimize multimodal perception models (CV and DSP) to handle unique signature detection tasks, such as identifying drones or aircraft from audio and video feeds and high and low altitude navigation;
- Develop and deploy scalable real-time vision systems using both classical computer vision and deep learning.
- Support and lead dataset curation and synthetic data generation pipelines for aerial and ground-based perception systems.
- Integrate perception models on embedded/edge hardware (Jetson, RPi5, etc.) with emphasis on latency, robustness, and throughput.
- Collaborate closely with hardware and prototyping teams to ensure synchronized sensor system operations across multi-camera and acoustic arrays.
- Contribute to the overall perception architecture, help standardize modular pipelines, and enable extensibility for future operational growth.
Required Qualifications
- Bachelor’s or Master’s degree in Computer Science, Electrical Engineering, or a related field.
- 8+ years of progressive experience in machine learning and computer vision roles.
- Strong Python development skills and solid C++ understanding; expertise in ML frameworks such as PyTorch, Detectron2, OpenMMLab, Ultralytics, etc.
- Proven track record in video analytics, UAV/drone imagery, acoustic signal processing, and applied 2D/3D CV (e.g., 3D object detection, segmentation, keypoint estimation).
- Experience in designing real-time CV pipelines, multi-sensor fusion, and implementing robust data labeling and active learning systems.
- Practical experience deploying perception solutions in both cloud and edge environments, including GPU-constrained platforms.
Nice to Have
- Familiarity with regulated environments (defense and medical), including model validation and performance benchmarking.
- Leadership experience in mentoring engineers, managing annotation teams, and driving end-to-end computer vision pipelines from prototyping to deployment.