Senior Applied Research Engineer

Summary

Join Spexi, a drone technology company, as a Senior Applied Research Engineer to lead the development of experimental algorithms and prototype systems for geospatial imagery. You will bridge early-stage research and production, delivering high-quality code for next-generation products. Responsibilities include developing novel algorithms in computer vision and geospatial processing, improving imagery accuracy using photogrammetric principles, investigating methods for geometric and semantic understanding of imagery, working with coordinate reference systems, writing performant code in C++ and Python, designing scalable pipelines for handling spatially indexed data, staying current on 3D reconstruction advances, and collaborating with product and engineering teams. This role requires 8+ years of experience in applied research or algorithm development, a Master's degree (Ph.D. preferred), strong proficiency in C++ and Python, and deep understanding of multi-view geometry and georeferenced data. The ideal candidate will also have experience with bundle adjustment, spatial databases, and cloud environments.

Requirements

• 8+ years of experience in applied research, software engineering, or algorithm development in computer vision, geospatial systems, or related domains
• Master’s degree in Computer Science, Imaging Science, Remote Sensing, or a related field. Ph.D. preferred
• Strong proficiency in C++ and Python, including libraries such as OpenCV, GDAL, and NumPy
• Deep understanding of multi-view geometry, 3D reconstruction, camera models, coordinate reference systems, projection transformation and georeferenced data
• Experience with bundle adjustment, aerial triangulation, and optimization libraries such as Ceres Solver
• Experience with spatial databases and tools such as PostGIS, and designing data structures for large-scale spatial indexing and querying
• Exposure to object detection (e.g., YOLO), semantic segmentation (e.g., SAM), and modern ML approaches to feature extraction
• Comfort working with large datasets in cloud environments (AWS) and optimizing prototypes for scalability

Responsibilities

• Develop and prototype novel algorithms in computer vision, 3D geometry, and geospatial processing that enable automated understanding, alignment, and retrieval of aerial imagery and its derivatives
• Design approaches to improve spatial accuracy of imagery using photogrammetric principles, including bundle adjustment and camera pose estimation, with limited or no ground control data
• Investigate methods for geometric and semantic understanding of imagery, including feature extraction, image segmentation, and object-level matching across views
• Work with coordinate reference systems, projection transforms and georectified imagery
• Write performant and maintainable code in C++ and Python to support prototyping and handoff to production engineering teams
• Design prototypes enabled by scalable pipelines for handling spatially indexed data, using tools like PostGIS and cloud-native services in AWS
• Stay current on advances in 3D reconstruction (e.g., multi-view stereo, SLAM, radiance fields) and apply relevant techniques to aerial data workflows
• Collaborate cross-functionally with product and engineering teams to evaluate feasibility, scope impact, and translate concepts into concrete implementation plans

Preferred Qualifications

• Experience with point cloud processing and tools such as PDAL
• Familiarity with radiance field modeling, NeRFs, or generative scene reconstruction techniques
• Familiarity with LLMs and Generative AI for imagery
• Previous work translating research into production-quality solutions in fast-paced or startup environments

Benefits

• Remote-friendly environment (with a hub in Vancouver, Canada)
• Flexible hours
• Medical, dental, and vision health benefits