Principal Software Engineer (ai / Agentic Developer Productivity)

Microsoft Microsoft · Big Tech · Redmond, WA +1 · Software Engineering

Principal Software Engineer to build and scale the agentic AI platform for developer productivity at Microsoft, focusing on orchestration of AI models and agentic workflows for internal developers.

What you'd actually do

  1. Own and deliver complete features across the development lifecycle, including design, architecture, implementation, testability, debugging, shipping, and servicing.
  2. Use AI in your daily coding and advocate to the team and customers an evolving state-of-the-art of how best to do this.
  3. Demonstrate resiliency to experiment and try multiple AI approaches that may not work — because no one has tried them before.
  4. Provide mentorship and coaching to engineers in, and beyond, your team.
  5. Write and review clean, well-thought-out code with an emphasis on quality, performance, simplicity, durability, scalability, and maintainability.

Skills

Required

  • Bachelor's Degree in Computer Science or related technical field AND 6+ years technical engineering experience with coding in languages including, but not limited to, C, C++, C#, Java, JavaScript, or Python OR equivalent experience.

Nice to have

  • Prior experience building with Agent SDKs, large language models (LLMs), prompt engineering, or AI orchestration frameworks (e.g., LangChain, Semantic Kernel, AutoGen, LlamaIndex, or similar).
  • Familiarity with agentic AI development patterns — multi-step reasoning, tool/function calling, retrieval-augmented generation (RAG), and human-in-the-loop workflows.
  • Proven experience as a Software Engineer delivering large-scale, production-quality systems.
  • Solid software engineering fundamentals: system design, algorithms, testing, debugging, and code review.
  • Demonstrated ability to lead technical direction and mentor peers in a collaborative team environment.
  • Comfort working in ambiguous, fast-moving problem spaces where best practices are still being established.
  • Hands-on experience with AI-powered developer tooling and coding assistants, such as: GitHub Copilot and Copilot Workspace, Claude Code, OpenAI Codex / ChatGPT, Cursor.
  • Prior experience using agentic plugins, agents, skills, hooks, etc.
  • Experience dogfooding, evaluating, and forming opinions on emerging AI developer tools, and translating those insights into team-wide best practices.
  • Background in developer tooling, build systems, CI/CD pipelines, or engineering systems at scale.

What the JD emphasized

  • agentic AI development
  • shipping with it
  • agentic platform
  • orchestrating the latest AI models
  • agentic development
  • AI-Powered developer efficiency
  • Agent SDKs
  • large language models (LLMs)
  • prompt engineering
  • AI orchestration frameworks
  • agentic AI development patterns
  • multi-step reasoning
  • tool/function calling
  • retrieval-augmented generation (RAG)
  • human-in-the-loop workflows
  • AI-powered developer tooling
  • coding assistants
  • agentic plugins
  • agents
  • skills
  • hooks
  • dogfooding
  • evaluating
  • forming opinions on emerging AI developer tools
  • translating those insights into team-wide best practices

Other signals

  • building the agentic platform itself
  • orchestrating the latest AI models
  • company-wide developer productivity
  • large-scale codebase refactoring
  • automated package integrations
  • deep architectural Q&A over codebases
  • human-in-the-loop directed code changes