In short
For our client, a fast-growing AI company building realtime voice and LLM technology, we are looking for a Senior Machine Learning Engineer with strong expertise in ML inference…
PYTHONNODEKUBERNETESLLMMACHINE LEARNING
📋 About the job
Senior Machine Learning Engineer, Inference
Job Summary
For our client, a fast-growing AI company building realtime voice and LLM technology, we are looking for a Senior Machine Learning Engineer with strong expertise in ML inference optimization to join its inference team.
In this role, you will own inference optimization end-to-end – taking models from research to production at consumer scale and solving latency and throughput problems for realtime AI products used by hundreds of millions of people.
Our client is a well-funded, venture-backed AI research lab and realtime inference provider with a small, senior engineering team.
Team
Inference : The team owns how our client’s models run in production – inference engines, GPU efficiency, model acceleration and the distributed serving infrastructure behind realtime voice and LLM products.
This is an opportunity for senior engineers who want to work deep inside modern inference engines on problems at real consumer scale.
Responsibilities
🔹Own inference optimization end-to-end, from research models to production deployment;
🔹Optimize latency and throughput of realtime voice and LLM inference at consumer scale;
🔹Work inside inference engines such as vLLM, TRT-LLM and SGLang to improve performance;
🔹Apply model acceleration techniques – quantization, distillation, speculative decoding, caching, continuous batching and paged attention;
🔹Write high-performance code in C++, CUDA, Rust or optimized Python for GPU workloads;
🔹Build and scale distributed multi-GPU and multi-node serving with Kubernetes, Ray and custom load balancing for high concurrency;
🔹Collaborate with research and product teams to bring new models into production.
Requirements
🔹5+ years of experience as an ML inference or systems engineer, at senior level (Google L5 / Meta E5 or equivalent);
🔹Hands-on ML inference optimization inside vLLM or TRT-LLM (engine-level work, not only deployment);
🔹Experience building backend/ML systems and serving models in production;
🔹Systems programming and GPU optimization skills: C++, CUDA, Rust or optimized Python;
🔹Experience with model acceleration: quantization, distillation, speculative decoding, caching, continuous batching, paged attention;
🔹Distributed systems experience: Kubernetes, Ray, load balancing, multi-GPU/multi-node inference, high concurrency;
🔹Based in the US, Canada, UK, Germany, Switzerland or Serbia;