CareerPlanSign in

Senior Applied Scientist - Copilot & Agents

United Kingdom, London, London💼 Full-time🗓 2026-07-24 → 2026-09-26

Core

Develop scientific approaches to evaluate and improve LLM reasoning, grounding, retrieval, and agentic behaviors through benchmarking, experimentation, and telemetry analysis.

Role type

Senior Applied Scientist (LLM & Agents)

Builds

Scalable, production-ready systems for LLMs, reasoning, and agentic workflows used by millions of customers.

Domain

Artificial Intelligence, Large Language Models, Machine Learning, Information Retrieval

Deliverable

production ML models

Required skills

LLM evaluation methodologies, offline benchmarking, online experimentation, telemetry analysis, agentic workflow design, multi-step reasoning, RAG, tool orchestration, reinforcement learning, model failure analysis, technical design review

Preferred skills

Ph.D. in CS/Math/Physics/Statistics, publication record in IR/ML/NLP, experience with recommendation systems

Technologies

Large Language Models, Reinforcement Learning, Retrieval-Augmented Generation (RAG)

Responsibilities

Define and execute rigorous evaluation methodologies and offline benchmarks, drive complex technical investigations to improve model quality and latency, partner with engineering to translate research into production systems, analyze model behavior and failure modes to identify improvements, develop approaches for agentic workflows and RL-based optimization, review technical designs for scientific rigor

Seniority

Senior, hands-on IC

Sourced via microsoft · Listed on CareerPlan, which tracks 70,000+ jobs from 20+ sources.