AI Computer Vision Lead

Quantum Systems next-generation unmanned solutions across air, land, and sea. From advanced aerial drones to underwater and ground platforms, our systems combine cutting‑edge hardware with AI‑driven software for seamless control and real‑time data insights. Trusted by defense, security, and public sector clients worldwide, we redefine mission versatility and performance.

Quantum-Systems is a strong supporter of Ukraine and has its local Ukrainian office since 2022.

The role is based in Kyiv and we expect you to work full‑time in the office.

• Lead team of ML Engineers
• Design, develop, and optimize real‑time object detection and tracking models for aerial video from UAVs (EO/IR cameras)
• Implement inference pipelines optimized for edge hardware (e.g., NVIDIA Jetson, Orin)
• Conduct model training, testing, benchmarking, and validation in real flight environments
• Integrate AI models into onboard systems
• Collaborate with other software and hardware teams to ensure robust end‑to‑end performance
• Research and apply state‑of‑the‑art deep learning methods in computer vision

• 8+ years of experience in AI/ML, with a strong focus on computer vision
• Strong leadership experience
• Experience building ML infrastructure and training pipelines
• Proven experience in object detection/tracking using models like YOLO, SSD, or custom CNNs
• Proficiency in Python, C++, PyTorch or Tensor Flow
• Experience with real‑time video processing and optimization techniques (e.g., TensorRT, ONNX, pruning, quantization)
• Solid understanding of data annotation, augmentation, and training workflows
• Familiarity with embedded/edge AI deployments (e.g., Jetson Xavier, Orin)
• Experience working with aerial/surveillance imagery or geospatial data is a strong plus
• Bachelor’s or Master’s degree in Computer Science, Robotics, or related field

• Knowledge of GStreamer, OpenCV, or similar real‑time streaming frameworks
• Familiarity with UAVs, drone flight control, or defense technologies
• Understanding of object re‑identification or multi‑target tracking