Research Engineer - Inference
ElevenLabs · Remote · United Kingdom · United States · Poland · Bulgaria ·
COMPENSATION
€•• – ••k
Listed on the employer's job board (Ashby)
- Fully remote (Global)
- AI Inference Optimization
- GPU & CUDA Programming
- Real-time Serving Systems
- No Degree Required
Join the research team at ElevenLabs to bridge the gap between experimental models and scalable products. You will own the systems that serve AI voice and chat agents to millions of users, ensuring high reliability and speed.
The work involves optimizing inference across the stack, from custom kernels to orchestration, enabling researchers to ship new capabilities quickly and safely.
What you'll do
- Deploy state-of-the-art models from research checkpoints to production infrastructure
- Optimize inference latency, throughput, and cost using quantization and distillation
- Build high-performance serving systems for real-time streaming workloads
- Implement KV-cache optimization, batching strategies, and custom kernels
- Create tooling to help researchers ship models with confidence in performance
What you'll need
- Experience deploying ML models in production for latency-sensitive apps
- Strong skills in GPU programming and inference optimization
- Proficiency with CUDA, Triton, TensorRT, vLLM, or SGLang
- Ability to profile and eliminate bottlenecks across the serving stack
- Capacity to build tooling for measuring system performance
Stack
- CUDA
- Triton
- TensorRT
- vLLM
- SGLang
How you'll work
- Fully remote with global eligibility
- Optional office access in London, New York, San Francisco, Warsaw
- High-velocity environment with lean autonomous teams
- Minimal bureaucracy and focus on impact over titles
What you get
- Annual discretionary stipend for professional development
- Annual discretionary stipend for social travel with colleagues
- Yearly company offsite in locations like Croatia or Italy
- Monthly co-working stipend for those not near main hubs
- Opportunity to define the trajectory of AI technology