Software Engineer (Computer Vision / Image Processing)

About this role

• Implement and re-implement expensive imaging-pipeline stages in performant C++, including bundle adjustment, orthorectification, and image format handling, and expose them cleanly to Python
• Drive down IR Data Product pipeline latency by replacing expensive stages with optimized implementations (e.g. GPU-based orthorectification), and extend the same approach to other imaging data product and missions
• Work with the image data products team to identify the highest value stages to port from Python to C++/GPU and drive the migrations to completion and success
• Contribute to high-performance reference algorithms and benchmark them
• Partner with the HPC GPU specialists where workloads benefit from GPU acceleration
• Own verification and validation, including unit, integration, and performance regression tests in CI
• Produce clear designs, trade studies, and interface documentation

### Benefits

• 100% health, dental, and vision coverage for employees; 50% coverage for dependents (Concierge Medical offered through One Medical where available)
• Parental leave for primary and secondary caregivers
• Generous time off policy
• 401(k) plan through Betterment and equity management through Carta
• Fully stocked kitchen at our HQ and we cater lunch 3 days per week
• Dog-friendly office- Ability and willingness to obtain and maintain a U.S. security clearance. Active clearance is a plus
• Bachelor’s or Master’s degree in Computer Science, Computer Engineering, Electrical Engineering, Applied Math, or a related technical field
• Expert-level C++, with deep familiarity with memory management, concurrency, and performance-oriented programming
• Strong written and verbal communication and ability to lead and conclude cross-functional technical discussions
• 3+ years of professional experience developing performance-critical C++ software in production
• Hands-on experience with bundle adjustment, orthorectification, and common image and geospatial formats
• Demonstrated exper...