Principal Applied Scientist

** Overview*
* Microsoft Cloud Operations and Innovation (CO+I) underpins Microsoft's global cloud infrastructure, driving the innovation, planning, design, construction, and operation of one of the largest data center fleets in the world. We are seeking an exceptional  
** Principal Applied Scientist
** to join the  
** Data Center Applied AI
** team. In this role, you will play a pivotal role in advancing and integrating  
** cutting‑edge, multimodal agentic AI systems
** into core tools and operational workflows that power Microsoft's data centers. Your work will drive operational efficiency at scale and help advance Microsoft's mission to empower every person and every organization on the planet to achieve more through intelligent, multimodal AI agents.

As a  
** Principal Applied Scientist
** t, you will  
** bridge state‑of‑the‑art AI research with production‑grade engineering** , delivering agent architectures that are  
** reliable, secure, scalable, observable, and measurable** . You will collaborate closely with Business and Engineering to build and deploy agentic capabilities across workflow automation, information retrieval, retrieval‑augmented generation (RAG), tool and function calling, long‑horizon task execution, and multi‑agent orchestration.

This role blends deep AI and applied science expertise with strong engineering judgment, operational rigor, and a bias for action. Success requires end‑to‑end ownership, a growth mindset, and strong customer empathy. Join us to shape the future of agentic AI for data center operations at global scale.

Microsoft's mission is to empower every person and every organization on the planet to achieve more. We work with a growth mindset, innovate to empower others, and collaborate to realize shared goals-guided by our values of respect, integrity, and accountability to foster an inclusive culture where everyone can thrive.

** Responsibilities*
* +  
** Advance Applied AI Research to Improve Quality and Reliability*
* + Apply deep expertise in  
** Generative AI, deep learning, NLP, and multimodal models
** to translate cutting-edge research into  
** high-impact, production-ready AI solutions** .

+ Design and execute experiments that measurably improve  
** agent planning, memory, grounding, reasoning, and long‑horizon task completion** .

+ Perform  
** lightweight fine‑tuning**  (e.g., LoRA and related techniques) on multimodal and large multimodal language models-to improve entity recognition, reasoning accuracy, and response quality.

+ Implement state‑of‑the‑art approaches using  
** foundation models, advanced prompt engineering, RAG, knowledge graphs, and multi‑agent architectures** , complemented by classical ML techniques where appropriate.

+  
** Build and Ship Scalable Agentic AI Systems*
* + Architect and implement  
** end‑to‑end agent workflows
** that decompose user intent into executable plans, intelligently select and invoke tools, and recover gracefully from errors and edge cases.

+ Rapidly prototype solutions and partner with engineering teams to drive  
** production deployment** , including debugging live systems and building  
** AIOps workflows** .

+ Use data and telemetry to identify AI quality gaps, generate insights, and deliver  
** proofs of concept
** that apply research innovations to real‑world operational challenges.

+ Design robust  
** multi‑step reasoning and tool‑use strategies** , including function calling, code execution, APIs, and secure connectors, with strong safety and reliability guardrails.

+ Drive production excellence across  
** latency, reliability, cost efficiency, observability, monitoring, and safe fallback behaviors** .

+  
** Own Evaluation, Metrics, and Iteration Loops*
* + Define task‑level evaluation metrics for agentic…