Computer Vision and Machine Learning Engineer

About the role

Join a dedicated R&D team focused on advancing MilTech products that have proven their worth in numerous missions. This role provides the opportunity to contribute to existing solutions, explore new methodologies, and see your work validated and scaled. You will own the implementation and refinement of machine learning and deep learning algorithms specifically for Last Mile Targeting, placing strong emphasis on tracking small objects in challenging conditions. The position involves developing and enhancing training pipelines for object detection and segmentation models while conducting rigorous data research and preparation. You will analyze data distributions, clean noisy or incorrect samples, and set up training balancing strategies to ensure robust model performance. A key responsibility is to optimize and compress models for ARM edge devices to meet real-world deployment constraints. You will stay current with research and approaches, analyze publications, and prototype novel solutions to improve computer vision and machine learning algorithms. The role also requires applying both classical and deep learning methods to object tracking and maintaining a proactive attitude towards continuous learning and self-improvement.

Key facts

What you'll do

• Implement and refine machine learning and deep learning algorithms for Last Mile Targeting, with a focus on tracking small objects.
• Develop and enhance training pipelines for object detection and segmentation models.
• Conduct data research and preparation, including analyzing data distributions, cleaning noisy or incorrect samples, and setting up training balancing strategies.
• Optimize and compress models for ARM edge devices.
• Stay current with research and approaches, analyze publications, and prototype novel solutions to improve computer vision and machine learning algorithms.
• Apply classical and deep learning methods to object tracking.
• Maintain a proactive approach to continuous learning and self-improvement.
• Collaborate with the R&D team to integrate developed algorithms into existing MilTech solutions.
• Translate research findings into production-ready modules that meet strict performance and reliability standards.
• Contribute to the full lifecycle of computer vision algorithm development, from dataset creation to deployment in the field.
• Utilize advanced model optimization techniques such as pruning, quantization (PTQ/QAT), knowledge distillation, and tensor/low-rank decomposition.
• Evaluate and compare the effectiveness of different methodologies to solve complex vision problems.
• Ensure that developed solutions are robust, scalable, and maintainable for operational use.
• Document processes and results to support knowledge sharing and future development efforts.

Requirements

• Minimum of 2 years of hands-on experience developing computer vision and machine learning algorithms for tasks such as detection, segmentation, classification, and tracking.
• Experience with training models in low-data scenarios, utilizing techniques like transfer learning and fine-tuning.
• Strong foundation in mathematics, including linear algebra, probability, and statistics, with a deep understanding of computer vision and machine learning principles.
• Practical experience with PyTorch, OpenCV, and ncnn.
• Experience in the full lifecycle of computer vision algorithm development, from dataset creation to deployment.
• Experience with model optimization techniques such as pruning, quantization (PTQ/QAT), knowledge distillation, and tensor/low-rank decomposition.
• Understanding of current trends and approaches in computer vision, machine learning, and deep learning.
• Experience applying both classical and deep learning methods to object tracking.
• A proactive attitude towards continuous learning and self-improvement.
• Strong problem-solving skills and the ability to work independently on complex technical challenges.
• Excellent written and verbal communication skills in English.

Nice to have

• Experience with Ultralytics and OpenVINO, along with related libraries.
• Product development mindset, with skills in rapid prototyping and creating Minimum Viable Products.
• Experience working with synthetic data and Sim2Real approaches.
• Understanding of epipolar geometry, camera calibration, and image stabilization.
• Prior experience with drones or robotics.

Practical notes

• Compensation is competitive and aligned with market rates, including performance-based bonuses and incentives.
• Military reservist status can be facilitated with current documentation.
• Relocation assistance is available for candidates moving from other cities or countries.
• Employees receive 24 calendar days of vacation annually, plus an additional day for their birthday.
• Flexible work arrangements include a flexible schedule and up to 10 remote workdays per year.
• Comprehensive health insurance from a leading provider is available after the probationary period.
• Professional development is supported through compensation for training, courses, and access to a corporate library.
• Mental health support is provided through company-paid psychologist consultations.
• Legal employment is structured under Diia.City and labor legislation, ensuring stable working conditions.