inference systems · language models
ABOUT THE ROLE
Join a small AI infrastructure team building efficient inference systems for open-source language models. You will optimize models and serving infrastructure across diverse hardware, helping improve performance, reliability, and cost per token.
WHAT YOU'LL DO
- Ship support for new open-source models, tuning them for latency and throughput.
- Optimize inference serving with techniques such as batching, KV caching, speculative decoding, and quantization.
- Write and tune kernels using CUDA, HIP, and Triton across a range of accelerator platforms.
- Design, deploy, and operate heterogeneous clusters across infrastructure vendors.
- Run production inference across a mixed hardware fleet, improving reliability, observability, and cost efficiency.
- Work with customers to understand technical needs and solve infrastructure problems.
WHAT WE'RE LOOKING FOR
- Typically 1 to 6 years of experience in backend, infrastructure, systems, or low-level engineering, including experience at a technology company or early-stage startup.
- Strong computer science fundamentals, including computer architecture, memory hierarchy, and operating system internals.
- Experience owning a challenging system end to end, notable technical achievements, or evidence of rapid growth.
- Systems engineering experience beyond application-level or full-stack development, with a non-research-focused background.
- Familiarity with CUDA, Triton, or GPU programming; experience with HIP, TPU, or AWS Trainium is also relevant.
- Knowledge of inference optimization technique