Applied Research - RL & Agents

Prime Intellect•San Francisco, CA

2d•$150,000 - $300,000•Hybrid

About The Position

Prime Intellect is building the open superintelligence stack, providing infrastructure for AI labs. Their platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment for post-training at frontier scale, including SFT, RL, tool use, agent workflows, and continuously improving production models. They are developing open frontier AI, including open-source models for long-horizon tasks and the platform their research team uses. Prime Intellect has raised $150M from prominent investors and individuals in the AI and infrastructure space. They are seeking individuals passionate about building at the intersection of frontier research, real infrastructure, and go-to-market for a nascent category.

Requirements

Strong background in machine learning engineering, with experience in post-training, RL, or large-scale model alignment.
Experience with agent frameworks and tooling (e.g. DSPy, LangGraph, MCP, Stagehand).
Familiarity with distributed training/inference frameworks (e.g., vLLM, sglang, Accelerate, Ray, Torch).
Track record of research contributions (publications, open-source contributions, benchmarks) in ML/RL.
Passion for advancing the state-of-the-art in reasoning and building practical, agentic AI systems.
Strong technical writing abilities (documentation, blogs, papers) and research taste.
Eagerness to drive collaborations with external partners and engage with the broader open-source community.

Nice To Haves

Experience with web programming (React, TypeScript, Next.js).
Experience running LLM evaluations and/or synthetic data generation.
Experience deploying containerized systems at scale (Docker, Kubernetes, Terraform).

Responsibilities

Designing and iterating on next-generation AI agents for real workloads like workflow automation, reasoning-intensive tasks, and decision-making at scale.
Developing systems and frameworks for reliable, efficient, and large-scale agent operation.
Translating ambiguous objectives into clear technical requirements for product and research priorities.
Rapidly designing and deploying agents, evaluations, and harnesses for real-world tasks to validate solutions.
Shaping the direction and feature set for verifiers, the Environments Hub, training services, and other research platform offerings.
Building high-quality examples, reference implementations, and “recipes” to facilitate extension of the stack.
Prototyping agents and evaluation harnesses tailored to real-world use cases and external systems.
Pairing with technical end-users to design environments, evaluations, and verifiers reflecting real workloads.
Designing and implementing novel RL and post-training methods (RLHF, RLVR, GRPO, etc.) for aligning large models with domain-specific tasks.
Building evaluations and harnesses to measure reasoning, robustness, and agentic behavior in real-world workflows.
Prototyping multi-agent and memory-augmented systems to expand capabilities for downstream applications.
Experimenting with post-training recipes to optimize downstream performance.
Rapidly prototyping and iterating on AI agents for automation, workflow orchestration, and decision-making.
Extending and integrating with agent frameworks to support evolving feature requests and performance requirements.
Architecting and maintaining distributed training/inference pipelines, ensuring scalability and cost efficiency.
Developing observability and monitoring (Prometheus, Grafana, tracing) for production deployment reliability and performance.