Staff AI Inference & Systems Engineer

Mixpeek; San Francisco (CA); On-site; USD 180,000 - 240,000; Full time

Job summary

Modal is building an infrastructure layer for AI workloads, covering training, deployment, observation, and inference. You will perform hands-on inference research, selecting high-impact bets with the research lead and owning them end-to-end to reduce cost per token and tail latency on customer workloads.

You will work with customers alongside Forward Deployed Engineers to deploy and tune models, while expanding collaborations with external labs and turning frontier techniques into usable

Qualifications

• Research-leaning or systems background in LLM inference with demonstrable work.
• Fluency across LLM serving stack from kernels to autoscaling.
• Track record of shipping research or systems used by others.
• Ability to independently take a research idea to results in the open.
• Willingness to work in NYC or San Francisco offices.

Responsibilities

• Own end-to-end inference research bets and investigate high-impact ideas.
• Train speculators against production traffic and learn from results.
• Collaborate with Forward Deployed Engineers to deploy and tune models.
• Collaborate with outside labs (e.g., ZLab, SGLang, Flash Attention 4 kernels).
• Turn frontier serving techniques into products with engineering teams.
• Help shape the research agenda and guide future work.

Skills

LLM inference

Research mindset

Shipping research

Tools

Quantization FP8/INT4

Autoscaling

Job description

Modal is building an infrastructure layer for AI workloads, covering training, deployment, observation, and inference. You will perform hands-on inference research, selecting high-impact bets with the research lead and owning them end-to-end to reduce cost per token and tail latency on customer workloads.

You will work with customers alongside Forward Deployed Engineers to deploy and tune models, while expanding collaborations with external labs and turning frontier techniques into usable