Principal AI SoC Runtime Software Architect

Join an early engineering team at an AI semiconductor startup focused on robotics and on-device computing. As the principal runtime architect you will define and build the software layer that coordinates sensor processing and AI inference across an embedded system-on-chip.

• Architect and build an embedded runtime that moves camera audio and IMU data through processing inference and robotics applications.
• Coordinate execution across CPUs ISPs DSPs and NPUs using DMA caching shared buffers and synchronization to deliver predictable low-latency pipelines.
• Define interfaces between the runtime Linux kernel drivers firmware compiler and software development kit including stable APIs observability and fault recovery.
• Debug and tune end-to-end performance across software and hardware measuring latency throughput bandwidth and power.
• Lead collaboration across runtime kernel driver multimedia and SDK engineering and help grow the team.

• 8 years of runtime software engineering or architecture experience including technical leadership on a runtime driver or multimedia project involving multiple engineers.
• Experience shipping a runtime on an embedded SoC with hands-on knowledge of computer architecture and data movement including DMA cache coherency and heterogeneous processors.
• Camera or imaging pipeline experience such as image processing ISP V4L2 or GStreamer.
• Strong C or C skills and experience with Embedded Linux and Linux kernel drivers.
• Experience building production runtime systems embedded middleware multimedia frameworks or performance-critical systems software.
• AI inference runtime execution provider or ML compiler integration experience is useful including ONNX Runtime TensorRT OpenVINO TensorFlow Lite QNN or TVM.
• A degree in electrical engineering computer engineering or a related field or equivalent practical experience.

Base salary: $200000 to $300000 annually.

On-site in Santa Clara California United States.