You and the process

With a person
Anna, in-house recruiter
9 years hiring engineers
30 minutes with a real recruiter
They read your CV with you, on a call, and say where the offers are being lost.
Didn’t find what you were looking for? Tell us what to build
Prime Intellect

Research Engineer - Inference

  • Hybrid

Salary

$150,000 - 350,000/ year

AI summary

For members

The whole posting in a few lines. Sign up to read it here and on every role you open.

Sign Up to Read

Description

The company is building an open superintelligence stack: the infrastructure that frontier AI labs usually build in-house, offered to AI teams.

Its platform, called Lab, combines compute, environments, evaluations, secure sandboxes, high-performance training and deployment for post-training at frontier scale, covering SFT, RL, tool use, agent workflows and production models that keep improving.

The company develops open frontier models trained end to end for long-horizon tasks such as autonomous research, and uses the same full-stack platform for its own research team.

It has raised $150M in total funding from Founders Fund, Radical Ventures and NVIDIA, along with individual investors including Andrej Karpathy and Dwarkesh Patel, and leaders and founders from many AI, infrastructure and enterprise companies.

The inference systems you build will support the company's RL research, scaling toward 1, 3T parameter models on NVL72 clusters, with a focus on rollout throughput and frequent model updates.

Responsibilities

  • Bring new model architectures into RL training by extending the vLLM integration in the RL training codebase, validating on small runs before scaling up.
  • Raise rollout throughput for long-horizon agents by tuning concurrency and KV-cache reuse as context lengths grow.
  • Improve multi-node MoE inference, including expert parallelism and the trade-offs of splitting prefill from decode.
  • Cut the overhead of trainer-to-inference weight updates using sharding-aware transfer and overlapped communication, while keeping rollout workers on the intended policy version.
  • Develop FP8 inference and assess lower precision, using trainer/sampler mismatch checks to find numerical regressions and measuring their effect on RL training.
  • Work with model researchers on architecture choices that improve inference efficiency, weighing model quality against execution cost on the target hardware.

Requirements

  • Hands-on experience building or optimizing GPU inference systems and a working knowledge of inference engine internals.
  • Strong Python skills and experience with PyTorch or a comparable ML framework.
  • Ability to profile end-to-end workloads and turn the results into verified performance gains.
  • Experience diagnosing numerical or integration problems in ML systems.
  • Contributions to vLLM or another inference engine.
  • CUDA or Triton kernel experience, particularly for low-precision inference.

Conditions

You can work remotely or in person at the San Francisco offices, so the arrangement is flexible.

Benefits

  • Cash compensation of $150-350k, plus equity incentives tied to the company's growth and your impact.
  • Health, dental and eye care, fully covered for you and your dependents.
  • Flexible PTO.
  • $200/month health and wellness benefit.
  • Visa sponsorship and relocation help for international candidates.
  • Professional development budget for courses and conferences.
  • Team off-sites every quarter, plus hackathons, conferences and learning opportunities.
  • Free meals at the San Francisco office when you work there.

Where you’d work

Part of the week in the office

You can work from

  • United States

Relocation offered

Visa sponsored

About the company

Prime Intellect

  • Industry: AI

Also hiring in San Francisco, United States

2 of their 11 open roles are remote

Your chances

Worth a look before you spend an evening tailoring a CV for it.

  • 23 checks run
  • 6 red flags

Still hiring?

14 checks

2 red flags

How crowded?

9 checks

4 red flags

Fits Me

How well does this role fit you?

Answer a few questions or drop your CV, and every role gets a fit score with the reasons, this one first.

  • Your field
  • Level
  • Stack
  • Work model
  • Salary floor
  • Must-haves
Details13 facts · Role, Location, Compensation, Employment, Company
Tech stack
  • Python
  • PyTorch
  • LangChain
  • CUDA
  • Databricks
Type
Full-time
Equity
Equity offered
Industry
AI
Specialty
ML
Region
United States
Show 7 more factsShow less

Role

Category
Data & Analytics
Specialty
ML
Tech stack
  • Python
  • PyTorch
  • LangChain
  • CUDA
  • Databricks

Location

Work model
Hybrid
Region
United States
Remote from
  • United States
Relocation
Offered
Visa sponsorship
Sponsored

Compensation

Salary
$150,000 - 350,000 / year
Pay period
Annual
Equity
Equity offered

Employment

Type
Full-time

Company

Industry
AI

Something wrong with this vacancy?

Similar vacancies

Share this vacancy

What's wrong with it?

The employer never sees who reported.

Reason