ML / CV Engineer (Thermal Detection & Tracking)

About the role

This position places you at the intersection of hardware innovation and real-world implementation within an R&D environment dedicated to advanced electronic solutions. You will own the development and refinement of algorithms that enable precise last-mile targeting through thermal imaging cameras, ensuring robust performance in operational scenarios. In this role, you will design, enhance, and optimize deep learning model training pipelines to support high-accuracy detection and tracking objectives. You will analyze, filter, and balance complex datasets, collaborating closely with the data team to ensure the integrity and utility of training information. Furthermore, you will optimize models for deployment on edge devices, applying techniques such as pruning, quantization, and distillation to meet strict computational constraints. You will also explore novel and unconventional methods to push the boundaries of algorithm performance and discover new pathways for innovation. Additionally, you will bring a product development mindset to every task, applying skills in rapid prototyping and MVP creation to deliver tangible results quickly. Your work will directly contribute to products that will see widespread implementation and shape technological advancements across target applications.

Key facts

What you'll do

• Develop and refine algorithms for last-mile targeting using thermal imaging cameras, focusing on accuracy and real-time constraints.
• Design, enhance, and optimize deep learning model training pipelines to streamline data ingestion, augmentation, and validation workflows.
• Analyze, filter, and balance datasets, collaborating closely with the data team to identify biases and improve data quality.
• Optimize models for edge devices through techniques like pruning, quantization, and distillation to ensure efficient inference and minimal latency.
• Explore novel and unconventional methods to improve algorithm performance, testing unconventional architectures and training strategies.
• Evaluate and integrate classical and deep learning approaches to tracking, selecting the best methodology for specific operational requirements.
• Maintain awareness of current trends in CV, ML, and DL, actively researching new papers, tools, and frameworks that can be applied to the problem space.
• Work hands-on with thermal cameras to understand imaging characteristics, limitations, and calibration procedures for better model development.
• Implement detection and segmentation algorithms that form the foundation for high-performance tracking solutions in complex environments.
• Contribute to the full lifecycle of CV algorithm development, from initial dataset creation and experimentation through to robust deployment and monitoring.
• Leverage experience with model optimization techniques to reduce model size and computational demand without sacrificing critical performance metrics.
• Utilize libraries such as PyTorch, Ultralytics, TFLite, OpenVino, NCNN, and OpenCV to build, train, and deploy efficient computer vision solutions.
• Document methodologies, experiments, and results to ensure reproducibility and knowledge sharing within the R&D team.
• Collaborate with cross-functional partners to translate business requirements into technical specifications and evaluation criteria.

Requirements

• Minimum 2 years of hands-on experience in developing CV and ML algorithms for detection and segmentation tasks in real-world scenarios.
• Strong mathematical foundation including linear algebra, statistics, and a deep understanding of CV/ML principles underlying modern algorithm design.
• Practical experience with libraries such as PyTorch, Ultralytics, TFLite, OpenVino, NCNN, and OpenCV for building and deploying models.
• Experience with the full lifecycle of CV algorithm development, from dataset creation and annotation through model training, validation, and deployment.
• Prior experience working with thermal cameras and understanding of their unique imaging characteristics and challenges.
• Familiarity with current trends in CV, ML, and DL, including recent advances in detection, segmentation, and tracking methodologies.
• Experience with both classical computer vision approaches and deep learning approaches to tracking, knowing when to apply each.
• A commitment to continuous learning and self-improvement, staying updated with new research, tools, and best practices in the field.
• Ability to write clean, efficient, and well-documented code that can be maintained and extended by other engineers.
• Willingness to iterate rapidly based on experimental results and feedback from colleagues and stakeholders.
• Strong problem-solving skills and the ability to debug complex issues in model behavior or data pipelines.
• Effective communication skills for collaborating with team members and explaining technical concepts to non-technical stakeholders.

Nice to have

• Product development mindset, with skills in rapid prototyping and MVP creation to accelerate the validation of new ideas.
• Experience with synthetic data and Sim2Real methodologies to bridge the gap between simulation and real-world performance.
• Familiarity with model optimization techniques like pruning, quantization, and knowledge distillation to improve efficiency.
• Understanding of classical computer vision approaches to provide baseline methods and fallback solutions.
• Knowledge of epipolar geometry, camera calibration, and visual image stabilization to enhance geometric reasoning.
• Experience with SLAM algorithms to support advanced tracking and mapping capabilities in dynamic environments.
• Background in drone or robotics applications, providing context for motion dynamics and sensor integration in mobile platforms.

Practical notes

• Employment through Diia.City, offering transparent collaboration and stable working conditions.
• 24 calendar days of annual leave, plus an additional day for your birthday.
• Flexible working hours and up to 10 days of remote work per year.
• Medical insurance from a leading provider, active after the probation period.
• Support for professional development, including course and training compensation, and access to a corporate library.
• Mental health support through company-paid psychologist consultations.
• Competitive compensation package with market-rate salary, bonuses, and performance-based incentives.
• Assistance with military registration for eligible candidates with current documentation.
• Relocation support, including financial aid for candidates moving from other cities or countries.
• Office space available in Kyiv.