Principal Applied AI Engineer

** Overview*
* ACES CØDE  is a Microsoft engineering organization focused on reducing operational demand across the Microsoft Cloud, including but not limited to the avoidable support cases, repeated troubleshooting, and manual interventions that arise when cloud products, platforms, or processes do not work as intended. Our mission is to engineer this demand out of the Microsoft Cloud by turning customer and operational signals into lasting product improvements, intelligent diagnostics, autonomous resolution systems, and AI-powered capabilities.

As cloud services grow in scale and complexity, we are moving beyond simply reacting to problems after they occur. We work to understand why operational demand exists, prevent it at the source where possible, resolve it autonomously when it does occur, and continuously learn from those signals to make our products and support experiences more reliable. We look for teammates who can navigate ambiguity, reason from data and customer signals, work across technical and organizational boundaries, and turn complex problems into durable, measurable outcomes.

Within this mission, the ACES CØDE engineering team is building domain-specific agents and services that are transforming Azure customer support from reactive, human-dependent workflows toward AI-first systems capable of resolving customer issues at Azure scale. We are seeking a Principal Applied AI Engineer to serve as a hands-on technical architect and senior developer for these production agentic AI services.

You will define and implement the architecture for next-generation AI agents, including LLM orchestration, multi-agent coordination, retrieval-augmented generation (RAG), tool integration, and evaluation systems. You will write production code, lead complex technical initiatives, establish engineering standards, and guide engineers across organizational boundaries through technical influence rather than people management.

This is a deeply technical individual contributor role with broad architectural scope and direct accountability for production outcomes. You will turn ambiguous customer and business problems into scalable systems, make critical design decisions, build foundational components, and help teams ship secure and reliable AI services that handle thousands of customer cases per month.

Every day, our customers stake their business and reputation on our cloud. You can help Azure Eng Ops provide our customers with the world-class cloud services they need to succeed.

Microsoft's mission is to empower every person and every organization on the planet to achieve more. As employees, we come together with a growth mindset, innovate to empower others, and collaborate to realize our shared goals. Each day, we build on our values of respect, integrity, and accountability to create a culture of inclusion where everyone can thrive at work and beyond.

** Responsibilities*
* + Architect and build production agentic AI systems:
Design and implement scalable AI agents and services, including LLM orchestration, multi-agent workflows, RAG pipelines, tool integrations, memory and state management, and human-in-the-loop experiences.

+ Lead through hands-on engineering:
Write, review, and maintain production-quality C# and Python code. Build critical components, reference implementations, prototypes, and reusable platform capabilities that accelerate delivery across the engineering organization.

+ Own end-to-end technical delivery:
Translate ambiguous customer support scenarios into clear technical requirements and take solutions from architecture and experimentation through implementation, deployment, operation, and continuous improvement.

+ Define technical architecture and engineering standards:
Establish architecture patterns, API and data contracts, code quality expectations, evaluation practices, deployment strategies, and observability standards for AI services. Drive technical consistency while allowing teams to evolve their solutions.

+ Advance evaluation-driven AI development:
Design and implement evaluation frameworks using rubrics, golden datasets, simulation, judge models, and offline and online evaluation. Define quality thresholds and graduation criteria for moving agents from experimentation through shadow mode and into autonomous operation.

+ Own production quality and reliability:
Ensure AI services meet demanding standards for correctness, availability, latency, scalability, cost, and safety. Diagnose complex production…