Sesame
ML Model Serving Engineer
$175K–$280K
San Francisco +2 · On-site · Full-time · Posted Nov 3, 2026
- Location
- San Francisco
New York
Bellevue
About the job
Optimize and extend the serving systems for Sesame’s language, speech, and vision models to deliver fast, reliable, cost-effective inference for a consumer product.
What you'll do
- Improve model-serving performance using techniques such as in-flight batching, caching, and custom kernels.
- Extend LLM serving frameworks and collaborate with infrastructure and training engineers on serving and model speed.
- Reduce model initialization times while maintaining model quality.
What they ask for
- Expertise in a differentiable array computing framework, preferably PyTorch.
- Expertise optimizing machine-learning models for reliable, high-throughput, low-latency serving.
- Significant systems programming experience.
- Significant performance engineering experience, including bottleneck analysis or low-level systems profiling.
Nice to have
- Familiarity with high-performance LLM serving, including VLLM or SGLang deployment and internals.
- Experience with public cloud platforms or deploying and scaling inference workloads using Kubernetes or Ray.
Pay
$175K – $280K (base salary)
Tools
PyTorch, VLLM, SGLang, GCP, AWS, Azure, Kubernetes, Ray
AI focus
Improve inference serving for LLM, speech, and vision models, including model and serving-framework optimization.
Model-written summary · the original posting is authoritative