- Experience
- 4+ yrs
- Salary
- —
- Openings
- 1
- Posted
- 4 days ago
- Work mode
- In office
- Resume
- Required to apply
Where you'll work
Sign in to tell us what does and doesn't work for you here — it sharpens every match we show you.
Job description
About the Role
Join a dynamic AI infrastructure company that is developing an innovative software stack to enable efficient execution of machine-learning workloads across a variety of hardware platforms. As a Compiler Engineer, you will take ownership of the graph optimization component of the AI compiler in a rapidly growing startup gaining significant commercial traction.
Key Responsibilities
- Design and implement optimization passes at the graph level to enhance the performance of a heterogeneous AI compiler.
- Focus on advanced techniques such as fusion, layout optimization, dead-code elimination, constant folding, and algebraic simplification to refine the compiler's operation.
- Adapt and evolve the compiler's intermediate representation to accommodate changes in model architectures and hardware targets.
- Utilize performance analysis tools to uncover opportunities for optimization and improve runtime efficiency.
- Collaborate closely with frontend and code-generation teams to ensure well-defined compiler interfaces and robust validation processes.
Candidate Profile
- Minimum of four years’ experience in compiler engineering with an emphasis on IR design or writing optimization passes.
- Proficient in C and C++ programming languages.
- Strong understanding of graph and tensor optimization strategies including fusion, tiling, and layout transformations.
- Experience or familiarity with compiler frameworks such as MLIR and XLA is considered a plus.
- Knowledge of PyTorch, JAX, TensorRT, GPUs, or other AI accelerators is advantageous.
- A passion for tackling challenging performance issues at the convergence of machine learning models, compilers, and hardware.