Senior Computer Vision Engineer (Geospatial AI)

🕵 We are looking for Computer Vision Engineer (Geospatial AI) Workload: Full-Time (40 hours/week) Duration: Project-based with strong potential for a long-term engagement About the Role We are looking for an experienced Senior Computer Vision Engineer with a strong background in Geospatial AI, Visual Localization, and Deep Learning to join a highly innovative project in the defense technology sector. You will work on challenging real-world problems involving video understanding, cross-view localization, image retrieval, geospatial reasoning, and large-scale visual search. This is an opportunity to contribute to a cutting-edge AI system that operates under demanding conditions and requires state-of-the-art computer vision solutions. Responsibilities Develop, maintain, and optimize the end-to-end Computer Vision pipeline Process and analyze video streams, including low-quality and degraded video inputs Design and implement cross-view feature matching and image retrieval systems Build, maintain, and optimize georeferenced tile databases and indexing systems Develop object georeferencing and spatial analytics capabilities Improve inference speed, scalability, and overall system performance Collaborate with cross-functional teams on system architecture and technical design Evaluate and integrate state-of-the-art computer vision and geospatial AI models Contribute to production-grade deployment and monitoring of AI services Must-Have Qualifications 4+ years of professional experience in Computer Vision, Deep Learning, or related fields Strong hands-on experience with: Geo-localization, place recognition, or image retrieval systems Foundation models for feature extraction such as DINOv2, CLIP, or equivalent Keypoint detection and matching frameworks including LightGlue, LoFTR, DISK, and SuperPoint Contrastive and metric learning techniques such as InfoNCE, Triplet Loss, and Hard Negative Mining Homography estimation and geometric verification Experience working with satellite imagery and geospatial datasets Solid understanding of: Coordinate reference systems (CRS) Geospatial data processing Tile indexing and georeferenced imagery GDAL and related geospatial tooling Experience with vector search technologies such as FAISS or other Approximate Nearest Neighbor (ANN) libraries Advanced Python development skills Production-level experience with PyTorch Experience building and optimizing video processing pipelines Strong Linux and Docker expertise Nice-to-Have Qualifications PhD in Computer Vision, Photogrammetry, Remote Sensing, Geomatics, Robotics, or a related discipline Publications or significant research contributions in: Cross-view localization Visual SLAM UAV navigation Geospatial AI Experience handling degraded, noisy, compressed, or low-resolution video streams Experience with: TensorRT ONNX Runtime Edge AI deployment CUDA optimization Background in defense, aerospace, autonomous systems, robotics, or related industries ⚠️ Important: Candidates who attach a CV to their application will be prioritized.