Description
Kog is hiring an AGCO engineer to build a system that explores, generates, compiles, verifies, and measures LLM execution optimizations across GPUs. The role covers compiler and IR design, optimization passes, code generation, search methods, correctness verification, GPU execution and profiling, LLM inference systems, and optimization loops. Candidates should have deep technical expertise in compilers, GPU programming, LLM inference, or related systems, with the ability to explain their technical work. The role is based in Paris and requires relocation.
