Applied Research - Forward-Deployed
$150k - $300k • San Francisco • FullTime
Posted 3mo ago
About the job
Prime Intellect is seeking a Forward-Deployed Research Engineer (FDRE) to act as the primary technical liaison between the company and its key clients, including AI companies, research labs, and enterprises. This role involves embedding with customers to understand their models, workflows, and objectives, and then translating these into practical training runs, environment designs, evaluation frameworks, and deployment strategies using the Lab platform. You will be instrumental in ensuring the platform's effectiveness for real-world applications and will collaborate with internal research, product, and infrastructure teams to integrate customer feedback and shape future development.
Responsibilities
- Embed with strategic customers to understand their agent architectures, failure modes, and product goals.
- Design and build custom RL environments, evaluation harnesses, and verifiers tailored to customer domains.
- Architect agent scaffolding, including tool use, multi-step reasoning, memory, and sandbox execution, for customer workflows.
- Configure and launch training runs on the Lab platform, iterating on reward functions, rollout strategies, and evaluation criteria.
- Serve as the end-to-end technical lead for customer engagements, from discovery to deployed models.
- Identify and codify repeatable patterns from customer engagements into reference implementations, templates, and documentation.
- Act as the voice of the customer internally, influencing the roadmap for Lab, verifiers, the Environments Hub, and training infrastructure.
- Build examples and 'recipes' to facilitate customer and contributor adoption of the stack.
- Contribute to technical content such as blog posts, tutorials, and case studies.
- Develop novel evaluation methodologies for agentic behavior, including multi-step reasoning and long-horizon task completion.
- Prototype and iterate on agent harnesses for real-world tasks like code generation and workflow automation.
- Experiment with reward design, rubric construction, and environment shaping to enhance training signal quality.
- Stay current on frontier agentic AI, evaluation methods, and post-training techniques, applying this knowledge to customer work.
Requirements
- Deep hands-on experience building, evaluating, or deploying LLM-based agents in the past 1-2 years.
- Strong intuition for evaluation design, including identifying key metrics, constructing rubrics, and assessing reward signal strength.
- Working understanding of RL and post-training concepts (GRPO, RLHF, reward modeling, SFT).
- Strong Python skills and familiarity with the modern AI stack (Hugging Face, inference engines, agent frameworks).
- Experience in a customer-facing or consulting-adjacent technical role, or as a technical founder.
- Excellent written and verbal communication skills.
- High agency and comfort with ambiguity, with the ability to scope problems, ship solutions, and iterate.
Benefits
- Cash Compensation Range of $150-300k
- Equity incentives
- Flexible Work