Requirements

• Experience working on ML compilers or other compiler engineering roles
• Familiarity with various parts of the ML compiler stack, including high-level ML frameworks (PyTorch, Jax, TensorFlow), graph compilers (Inductor, XLA, TVM), or kernel DSLs (Triton, Helion, cuTile)
• Understanding of compiler concepts such as static analysis, intermediate representations (SSA, CFGs), and instruction selection
• Fluency in English

Nice to Haves

• Interest or experience in ML compiler optimizations like algebraic rewriting, fusion, scheduling languages, loop-level optimization, and autotuning
• Knowledge of ML compiler frontends including graph capture, symbolic and dynamic shapes, control flow, and automatic differentiation
• Experience with GPU kernel programming and architecture, including Triton, cuTile, CUDA, and programming with tensor cores
• Expertise in ML inference techniques such as CUDA Graphs, KV caching, and batching
• Familiarity with ML training methods including parallelism strategies (Tensor, Sequence, Expert, Pipeline), FSDP, NCCL, and activation checkpointing
• Skills in GPU debugging and optimization tools such as CUDA-GDB, Nsight Systems, Nsight Compute
• Interest in formal methods like proof assistants, automated theorem provers, and agentic proof systems
• Background in programming languages and DSL design, including embedded DSLs, staging, metaprogramming, type systems, and functional programming

What You'll Be Doing

• Shaping the development of the ML compiler stack to enable rapid model development and deployment
• Improving existing ML compiler technologies and exploring opportunities to build custom compiler solutions
• Working in a collaborative environment focused on designing delightful APIs and software systems