09/13/2026 updated
100 % available
Computer Vision & Machine learning Engineer
Toulouse, France
Only remote
Master of Computer Science in Artificial Intelligence and Pattern RecognitionAbout me
Computer Vision Engineer specializing in real-time video processing and embedded deep learning. Over 4 years of experience on critical systems (aeronautics/aerospace) — C++/CUDA expert, Python, AI architectures in production. Dual expertise in high-performance embedded vision and generative AI.
Artificial IntelligenceAmazon Web ServicesData AnalysisComputer VisionProgram OptimizationNvidia CUDAImage ProcessingMachine LearningMathematical ModelingOpenCVOpenGLPerformance TuningSoftware ArchitectureEye TrackingTensorFlow
03/2025 – Present
Computer Vision Engineer — HEMERIA (aerospace sector)
Real-time video processing at 100 Hz for safety-critical vision systems — detection/tracking in under 10 ms, a 100x performance gain GPU vs. baseline CPU
Real-time architecture for high-resolution video streams at high framerates; camera acquisition integration via low-level protocols (RDMA, Camera Link).
CUDA optimization: 700x700 image processing at 100 FPS, critical algorithms (filtering, segmentation, tracking) optimized to under 1 ms.
OpenGL shaders for GPU rendering/processing; fine-grained GPU memory and CPU-GPU transfer optimization (zero-copy, pinned memory); Nsight profiling.
Deep learning models optimized for real-time inference; quantization, pruning, TensorRT for embedded constraints.
RDMA protocols for ultra-low-latency GPU-to-GPU data transfer; buffer/queue management to minimize latency.
10/2021 – 12/2024
Computer Vision Engineer — HINFACT (aeronautics sector)
Real-time eye-tracking and behavioral analysis at 60 Hz+ — accuracy under 1° in 90% of cases, under 5° in 95% of cases.
Deep learning models for eye tracking; software architecture for simultaneous processing of multiple data streams (video, physiological sensors, operational data).
Advanced optimization: expert-level multithreading, GPU acceleration (CUDA), robust design patterns.
Real-time video streaming with GPU acceleration (FFmpeg, OpenGL); infrared camera integration for variable lighting conditions.
Technologies: C++, Qt, OpenCV, TensorFlow, PyTorch, CUDA, FFmpeg, NIR cameras.
Additional achievements :
3D Vision: multi-camera calibration, dense 3D point clouds (OpenGL/PCL), object localization and regions of interest. (C++, Python, OpenCV, Intel RealSense)
Generative AI (2023): RAG system POC with chatbot interface, delivered in 2 weeks (team of 3) adopted and used in production. (LangChain, HuggingFace, AWS, Docker...)
Cross-functional Skills
Robust and maintainable software architecture, CPU/GPU optimization, Agile / V-model methodologies, technical documentation, Git/GitLab CI/CD.
Experience
Professional experience5 - 10 years
Freelancer since2025
Languages
EnglishFluentFrenchNative speaker