Research Engineer, Knowledge Foundations na Anthropic

Híbrido - San Francisco, CA

Candidatar-se
Ver mais vagas na Anthropic

The Knowledge Work team at Anthropic builds training environments and evaluations that enable Claude to search, retrieve, and reason over information at scale. As a Research Engineer on Knowledge, you will design and run experiments, build infrastructure, train models, and develop evaluations to improve Claude’s performance on knowledge‑intensive tasks. You will work across the stack—from data pipelines to model training and evaluation—partnering closely with researchers and other RL teams to ship capabilities that directly influence Claude’s behavior.

Salary

USD 350,000 - 850,000

Requirements

Skills

  • Python engineering experience
  • Experience designing, running, and analyzing ML experiments
  • Ability to work across the stack (data pipelines, model training, evaluation)
  • 5+ years of experience operating ML or distributed systems at scale
  • Comfort with ambiguity and prioritization
  • Clear written and verbal communication
  • Ability to work with teams across time zones
  • Hands‑on experience training, fine‑tuning, or doing RL on large language models
  • Experience building evaluations for LLMs, especially in knowledge‑intensive domains
  • Prior work in a research‑heavy environment (e.g., frontier AI lab, quant research firm, domain‑focused AI startup)
  • Published research on LLMs, RL, retrieval or related areas
  • Experience with distributed training systems
  • Comfort with long‑term ownership of a system and its operational health

Responsibilities

  • Design, build, and iterate on training environments and data pipelines to improve Claude’s reasoning over knowledge‑intensive tasks
  • Run end‑to‑end experiments: form hypotheses, build infrastructure, train models, analyze results, iterate
  • Develop evaluations that capture progress on search, retrieval, and reasoning quality
  • Identify failure modes in current model behavior and translate them into training signals
  • Collaborate with researchers across RL Data, post‑training, and product teams to align on priorities and ship improvements
  • Contribute to shared infrastructure and tooling that increases team velocity
  • Own a clean, canonical set of evaluation tools and processes for Knowledge Work capabilities
  • Build and automate observability, dashboards, and operational tooling for training environments and evaluation systems

Technologies

PythonML experimentsData pipelinesModel trainingEvaluation toolsDistributed training systemsRLLLM

Compartilhar vaga

Descubra se seu currículo está pronto para esta vaga

Veja como nossa IA pode otimizar seu currículo e aumentar suas chances de conseguir esta posição.