Member of Technical Staff (Machine Learning Engineer, Search & Agents)

About the role

Perplexity is looking for an experienced Machine Learning Engineer to advance AI systems that search, reason, and collaborate to solve complex problems. You will work across the full stack—from models and agent harnesses to search infrastructure—to turn experimental ideas into production‑ready systems.

Key responsibilities

• Push search and agent quality forward through improvements to models, training data, tools, and system design.
• Develop LLM post‑training methods, including reinforcement learning, to improve reasoning, search, tool use, and task completion.
• Train and evaluate multi‑agent systems, exploring coordination, information sharing, and division of work.
• Design and build agent harnesses that provide context management, execution environments, and orchestration for reliable multi‑step work.
• Improve retrieval and ranking models and the search interfaces agents use to find and assess information.
• Build datasets, reward signals, and evaluations that expose failures and guide improvements.
• Own experiments end‑to‑end, from hypothesis formulation to scalable training, deployment, and measurable gains.
• Collaborate with AI, Search, Infrastructure, Data, and Product teams to bring new capabilities into production.

Required profile

• Proven track record of building and shipping ML systems, especially in LLM post‑training, reinforcement learning, search/retrieval, or agent systems.
• Strong software engineering skills across model training, experimentation infrastructure, and production systems.
• Experience designing rigorous evaluations, diagnosing failures, and turning experimental results into practical improvements.
• Comfort with open‑ended problems that require both research judgment and hands‑on engineering.
• Ownership mindset, curiosity, and drive to see ideas through to working systems.

Required skills

• Machine learning engineering for large language models.
• Reinforcement learning and post‑training techniques.
• Search, retrieval, and ranking system development.
• Multi‑agent system training, coordination, and evaluation.
• Building scalable inference and distributed training infrastructure.