Jobs

Callosum is seeking an Inference Engine Development - Member of Technical Staff to build next-generation AI infrastructure that orchestrates across heterogeneous hardware. This role involves adapting and extending inference engines like vLLM and SGLang to be hardware-aware, optimizing scheduling, memory, and execution for diverse resources. You will design execution strategies such as parallelism and disaggregation to define heterogeneous inference at production scale. The ideal candidate has deep familiarity with inference serving frameworks, strong Python and C++/CUDA skills, and experience with parallelism strategies for large model serving.