Tomas P.

I help companies design, train, and deploy computer vision systems that solve real-world problems — not just research demos. I specialize in end-to-end computer vision solutions, taking ownership of everything from data and model performance to system reliability and deployment. What I help clients build Real-time computer vision systems (detection, tracking, segmentation) Custom deep learning models (CNNs, Transformers, U-Net variants) Medical image analysis (retinal imaging, segmentation, diagnostics support) Agricultural computer vision (crop analysis, monitoring, automation) Industrial & inspection systems (quality control, defect detection) Training pipelines & ML automation (data processing, reproducibility, CI-ready workflows) Why clients hire me 5+ years of software engineering experience, plus deep specialization in Computer Vision & ML Applied CV experience across multiple industries (medical, agriculture, industrial, AR research) Master’s degree in Computer Vision, thesis focused on retinal choroid segmentation Strong focus on production-quality code, documentation, and maintainability Comfortable working beyond the model: APIs, pipelines, backend integration, deployment Tech stack ML & CV: Python, PyTorch, TensorFlow, OpenCV Vision tasks: Detection, segmentation, tracking, image processing and analysis, stereo calibration Backend: Python, FastAPI, Java, Spring Boot, C# Frontend & tools: TypeScript, JavaScript, Qt Data: PostgreSQL, MySQL DevOps & MLOps: Docker, CI/CD Testing: Automated & end-to-end testing (WebdriverIO) How I work I don’t just deliver a model — I deliver a working system. I focus on: Clear requirements Measurable performance Clean, maintainable code Honest communication and ownership If you need a reliable computer vision engineer who can take a system from idea to production, let’s talk.