Lead AI Infrastructure Engineer
About this role
Avride needs to resolve pressing GPU inference bottlenecks across both real-time onboard autonomy systems and high-throughput offboard simulations. The incoming Lead AI Infrastructure Engineer will direct the foundational infrastructure layer that directly supports the company's autonomous vehicle software stack. This professional will initially optimize the GPU inference framework before expanding their ownership across broader machine learning pipelines.
Working closely with applied ML teams, the engineer shapes the company-wide architecture used for execution graphs and neural network runtimes. Success in this position requires a blend of deep systems knowledge and leadership capabilities to guide critical infrastructure decisions for autonomous technology.
What you will do
- Take ownership of the GPU inference framework with a primary focus on maximizing performance.
- Drive broader machine learning infrastructure responsibilities across various pipelines after initial onboarding.
- Collaborate directly with the applied ML team to align infrastructure with neural model architectures.
- Design and optimize execution graphs and runtimes used both onboard vehicles and in simulation environments.
What you need
- Bring five or more years of professional experience working with C++.
- Demonstrate a strong background in building robust infrastructure, including distributed systems.
- Possess practical experience working with the PyTorch framework.
- Show a solid understanding of underlying GPU architecture and operation.
- Demonstrate a proven ability to diagnose and resolve complex performance issues.
- Have programming experience within multi-threaded environments utilizing multiple processes, threads, timers, and interrupts.
Skills and tech stack
Questions candidates ask
Where is this job located and is remote work permitted?
This position is located in Austin, Texas. Remote work options are not available for this role.
Will the company provide visa sponsorship or relocation assistance?
The employer is not offering relocation or visa sponsorship. Candidates are required to be authorized to work in the U.S.
What will be the immediate focus for this role?
At first, you will take on the GPU inference framework, focusing heavily on performance. The biggest challenge is the effectiveness of GPU inference for both onboard applications and offboard cases.
Who will I work closely with on the engineering team?
You will work in close collaboration with the applied team responsible for defining neural model architectures. This partnership supports the execution graph building and optimization systems.