Senior / Lead Computer Vision Engineer

We are looking for a Senior/Lead Computer Vision Engineer to design and build high-performance perception systems for autonomous platforms, robotics, and advanced navigation operating under strict real-time and hardware constraints.

You will shape the architecture, implementation, and optimization of the entire vision pipeline - from geometric processing and object detection to tracking, sensor fusion, and integration with control systems. This role is ideal for someone who can independently solve hard technical problems and deliver reliable systems on embedded hardware.
You will work directly on defense projects for Ukraine, seeing a short path from code implementation to field testing.

What We Offer

• The chance to design mission-critical autonomous vision systems.
• Direct contribution to defense projects for Ukraine.
• Collaboration with top experts in vision, robotics, and embedded systems.
• Ownership of key architectural and technical decisions.
• Competitive compensation based on experience.
• Temporary military service exemption for work on defense-critical projects.

Requirements

Core Technical Skills

• 10+ years commercial software development; 5+ years in CV/ML or robotics.
• Degree in Computer Science, Math, Physics, or similar.
• Proven record of successful design and implementation of real-time CV or control algorithms.
• Strong in:
  • Object detection & tracking, Kalman filtering, sensor fusion.
  • Real-time control or optimal control algorithms.
  • Excellent C++14/17/20 and solid Python.
  • Experience with NVIDIA Jetson, ARM SoCs.
  • Low-latency video pipelines: GStreamer, DeepStream.
  • OpenCV, linear algebra, numerical methods.
  • Ability to work independently with minimal supervision.

Responsibilities

This role covers a range of responsibilities across computer vision, embedded systems, and robotics. You are not expected to cover all areas - the exact scope will depend on your experience and strengths.

• Architect and implement vision algorithms for autonomous platforms
• Optimize CV/ML models for real-time performance on constrained hardware
• Design and maintain low-latency camera pipelines (GStreamer, CUDA, DeepStream)
• Integrate vision with navigation, sensor fusion (IMU), and control systems
• Perform camera calibration, geometric vision, and sensor alignment
• Research and evaluate new algorithms to improve accuracy and robustness
• Participate in hardware/sensor selection and system architecture design
• Ensure robustness, safety, and performance of deployed systems
• Support field testing and system maintenance

Nice to Have

• Background in camera calibration, multi-view geometry, and 3D reconstruction
• Experience with SLAM, visual odometry, and multi-sensor fusion (camera, IMU, LiDAR)
• Practical knowledge of ROS/ROS2 and autonomous robotic systems, including drones, UGVs, or UAVs
• Expertise in PyTorch or TensorFlow, with additional experience in embedded ML optimization (TensorRT, ONNX, quantization, pruning)
• Familiarity with night-vision and infrared imaging systems
• Experience integrating platforms and protocols such as ArduPilot, PX4, MAVLink, or Betaflight
• Understanding of MLOps, deployment pipelines, and remote monitoring/telemetry solutions
• Exposure to mission-critical or functional safety environments

Language requirements:

English B1-B2 – Intermediate

Ukrainian C2 – Fluent

About the client:

This is a dynamic and rapidly developing company founded by experienced professionals. They offer exciting software engineering opportunities for Ukrainian engineers and work with clients both locally and internationally. They provide a full range of software development services, across multiple business verticals and a broad technological landscape.

Interview stages:

1. Interview with the agency recruiter and the client's representative, focusing on experience and organizational aspects.
2. Technical interview.

Information about the processing of your personal data is provided in our Privacy Policy, which is available online at Privacy Policy.