About the role
At River AI, our mission is to create personal AI owned and shaped by each individual. To achieve this, we are rewriting the entire stack from scratch: personal hardware for local inference, custom training infrastructure, next-generation UIs, and frontier deep learning research.
Who we are
We are scientists, engineers, and builders from the industry's top tech companies and AI labs. We bring a proven track record of scaling consumer systems for hundreds of millions of users and architecting the pre-training infrastructure behind today's frontier models.
About the Role
We are looking for exceptional AI compiler engineers to build the software bridge between the newest AI models and our high-performance custom silicon. You will create and build the compiler stack from PyTorch graphs all the way to optimized custom ISA assembly code. You will take ownership of kernel algorithms, intermediate representations, and even modify the ISA as necessary to achieve a flexible and high performance compiler stack. You will be collaborating both up and down the stack with AI researchers and modelers, as well as with performance engineers and silicon architects.
What You’ll Do
- Graph Lowering & Optimization: Design and implement compiler passes to lower PyTorch models into custom hardware, leveraging MLIR dialects and LLVM frameworks.
- Custom Backend Development: Develop and maintain the backend toolchain for our custom silicon, including instruction scheduling, register allocation, and hardware-specific code generation.
- Memory & Loop Transformations: Design sophisticated tiling and fusion strategies to maximize bandwidth utilization and minimize on-chip memory movement.
- Kernel Integration: Collaborate with software and hardware teams to integrate high-performance kernels (Triton/CUDA-like) into the automated compiler flow.
- Performance Profiling: Identify "compilation gaps" where the compiler fails to achieve peak hardware performance, and collaborate with the performance team for targeted optimizations to close those gaps.
- HW/SW Co-Design: Partner with the RTL and Architecture teams to change the custom ISA definitions.
Skills and Qualifications
Minimum Qualifications
- Bachelor’s degree in Electrical Engineering or Computer Engineering, and 5+ years practical industry experience working with advanced process nodes (7nm or below).
- Deep hands-on experience with MLIR or XLA for deep learning workloads.
- Expert-level understanding of PyTorch internals and how they interface with external backends.
- Proficiency in modern C/C++ for building robust, scalable, and high-performance compiler infrastructure.
- Advanced knowledge in Computer Architecture, especially the Programming Model, of at least one style of chip, including SoCs, CPUs, GPUs, or AI accelerators
- A highly collaborative mindset to push boundaries and co-design effectively with other engineers.
Preferred Qualifications: (We encourage you to apply even if you don't meet all of these)
- Hands-on experience in post-Silicon firmware and model update patches
- Experience defining and implementing custom dialects, lowering passes, and graph rewrites in an LLVM-based ecosystem.
- Knowledge of the tradeoffs between static and runtime environments, including JITs and ABIs
Logistics & Benefits
- Location: This role is based in Austin, Texas or Palo Alto, California.
- Compensation: Depending on background, skills, and experience, the expected annual salary range for this position is $200,000 - $420,000 USD, plus equity.
- Visa Sponsorship: We sponsor visas. We can't guarantee success for every candidate or role, but if you're the right fit, we're committed to working through the visa process.
- Benefits: River AI offers generous health, dental, and vision benefits, unlimited PTO, and relocation support as needed.