Senior Applied Scientist – Microsoft 365 Copilot  

Advance Retrieval Science Design and run experiments, define offline and online evaluation metrics, and develop scalable retrieval pipelines and models for enterprise-scale search systems. Areas of focus include: Semantic retrieval using late-interaction architectures such as ColBERT Dense retrieval and embedding model fine tuning Modern lexical retrieval approaches such as SPLADE Hybrid retrieval systems combining dense + sparse retrieval Query understanding and representation learning Multi-stage ranking and retrieval optimisation Retrieval-augmented generation (RAG) Personalization and contextual ranking Knowledge retrieval for agentic AI systems Reinforcement learning and reasoning-aware retrieval systems LLM-integrated retrieval architectures You will apply best practices in Responsible AI, Privacy-Preserving ML, and scalability for production-grade enterprise systems. Drive Product Innovation Partner with Engineering, PM and Design to translate product requirements and research advances into scalable and reliable retrieval infrastructure supporting Copilot Search, Chat and Agent experiences. Champion Customer Impact Deeply understand user retrieval pain points and enterprise grounding challenges, and develop solutions that materially improve relevance, answer quality, freshness and personalization. Lead and Mentor Provide technical leadership and mentorship to scientists and engineers working on retrieval, ranking and recommendation systems. Help establish best practices and contribute to the broader retrieval science strategy across CACore. Define Success Establish and evolve evaluation frameworks and success metrics for retrieval quality, grounding relevance, ranking effectiveness and downstream Copilot quality metrics. Stay Ahead Keep up with the latest advances in retrieval and ranking research, including developments in semantic retrieval, sparse retrieval, RAG systems and LLM-grounded search. Publishing at top-tier venues such as SIGIR, RecSys, WSDM, KDD, ACL and EMNLP is encouraged. Bachelor's Degree in Statistics, Econometrics, Computer Science, Electrical or Computer Engineering, or related field AND 4+ years related experience (e.g., statistics predictive analytics, research) OR Master's Degree in Statistics, Econometrics, Computer Science, Electrical or Computer Engineering, or related field AND 3+ years related experience (e.g., statistics, predictive analytics, research) OR Doctorate in Statistics, Econometrics, Computer Science, Electrical or Computer Engineering, or related field AND 1+ year(s) related experience (e.g., statistics, predictive analytics, research) OR equivalent experience. Strong hands-on experience developing retrieval or ranking systems at production scale. Demonstrated expertise in one or more of the following: Semantic retrieval Dense retrieval systems Embedding model training or fine tuning SPLADE or sparse retrieval methods Hybrid retrieval architectures Ranking systems for search or recommendation Large-scale information retrieval systems Experience developing ML systems in Python and modern ML frameworks such as PyTorch. Experience evaluating retrieval quality using offline metrics and/or online experimentation. Experience developing retrieval systems for RAG or agentic AI architectures. Publications in top-tier conferences such as SIGIR, RecSys, KDD, WWW, WSDM, ACL or EMNLP. Experience shipping retrieval systems integrated with LLM-based products. Familiarity with enterprise search, personalization and recommendation systems. Experience optimizing retrieval latency, scalability and serving infrastructure. Experience with reinforcement learning or retrieval-aware reasoning systems.