Modal
Member of Technical Staff - Research, Inference
$150K–$350K
New York +1 · Staff · Full-time · Posted Jul 5, 2026
- Location
- New York
San Francisco
About the job
Conduct hands-on research to improve LLM inference cost and tail latency, then work with customers and engineering to turn effective serving techniques into products.
What you'll do
- Own inference research from idea to result, including speculative decoding, quantization, KV-cache management, and autoscaling.
- Train custom speculators using production traffic and apply findings to target models.
- Collaborate with customers, research labs, and engineering to deploy, tune, and productize inference techniques.
What they ask for
- Research-leaning or systems background in LLM inference, with work to demonstrate.
- Fluency in the LLM serving stack, from kernels and quantization to schedulers and autoscaling.
- Record of shipping research or systems that others build on.
- Able to work in person in the NYC or San Francisco office.
Pay
$150K – $350K • Offers Equity
Tools
LLM, FP8, INT4, KV-cache, SGLang, Flash Attention 4
AI focus
Research and product development for LLM inference, including speculative decoding, quantization, memory management, and serving optimization.
Model-written summary · the original posting is authoritative