LLM Inference Engineer
Job Description
- Optimizing inference for the largest open-source model
- High-performance model serving frameworks such as TensorRT-LLM, vLLM, SGLang
- Distributed inference techniques (tensor/expert/pipeline parallelism, speculative decoding, KV cache management)
- Optimizing GPU utilization and latency for reinforcement learning
About Periodic Labs
We are an AI + physical sciences lab building state of the art models to make novel scientific discoveries. We are well funded and growing rapidly. Team members are owners who identity and solve problems without boundaries or bureaucracy. We eagerly learn new tools and new science to push forward our mission.
About the role
You will integrate, optimize, and operate large-scale inference systems to power AI scientific research. You will build and maintain high-performance serving infrastructure that delivers low-latency, high-throughput access to large language models across thousands of GPUs. You will work closely with researchers and engineers to integrate cutting-edge inference into large-scale reinforcement learning workloads. You will build tools and directly support frontier-scale experiments to make Periodic Labs the world’s best AI + science lab. You will make contributions to open-source LLM inference software.
You might thrive in this role if you have experience with:
- Optimizing inference for the largest open-source model
- High-performance model serving frameworks such as TensorRT-LLM, vLLM, SGLang
- Distributed inference techniques (tensor/expert/pipeline parallelism, speculative decoding, KV cache management)
- Optimizing GPU utilization and latency for reinforcement learning
