- Lead the design and implementation of robust and scalable C++ libraries and infrastructure to support use-cases as deep learning model execution on embedded devices, multi-sensor data collection, and system-level evaluation.
- Optimize the end-to-end runtime and latency of C++ vision inference library on embedded devices, often leveraging CUDA.
- Develop software to integrate the computer vision inference code into Anduril’s varied autonomous platforms including Sentry Tower, Lattice Ghost UAS, and more.
- Collaborate with hardware and and sensor teams to craft new platforms and prototype new perception systems.
- Write clean and reusable code
- Debug and root-cause perception related problems often spanning multiple devices and software systems in the customer environment.
- Bachelor’s degree in Computer Science, Electrical Engineering or related technical field
- 4+ years of industry experience
- 3 or more years C++ experience used in the last 4 year, including modern C++ features
- Proven track record of software development, including shipping one or more products
- Experience optimizing embedded software
- Able to both iterate quickly to produce a workable solution and think ahead for future improvements.
- Able to independently prioritize, scope work and collaborate across teams.
- Familiarity with machine learning, computer vision, and/or robotics.
- Experience with GPU programming
- Experience with OS internals and kernel development/debugging
- Experience with realtime sensor fusion (e.g. LiDAR, camera, radar)