Principal Research Software Engineer

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

Principal Research Software Engineer to provide technical leadership and direct technical contribution on the AI Agentic Core Team. The mission is to accelerate the path from research to product by building AI-driven systems, workflows, and platforms that help researchers and product teams move faster from exploration to real-world impact. This role involves collaborating with engineers, researchers, and product teams to build high-impact systems spanning early-stage prototypes through production-ready tools, services, and experiences, while modernizing how software is designed, built, evaluated, and shipped. The role requires designing, developing, and shipping systems that transition MSR concepts into production-quality tools, services, and product capabilities, owning the end-to-end engineering lifecycle. It also involves defining and implementing AI-driven processes that accelerate research-to-product pipelines using LLMs, agentic workflows, and modern developer tooling, including designing and integrating agentic AI frameworks and LLM-based pipelines, developing tool-use and function-calling architectures, and applying prompt design, RAG, and evaluation frameworks. Contributions to model experimentation and fine-tuning are also part of the role.

What you'd actually do

  1. Design, develop, and ship systems that transition Microsoft Research (MSR) concepts into production-quality tools, services, and product capabilities, owning the end-to-end engineering lifecycle from prototype to scalable, maintainable software.
  2. Build robust solutions spanning product experiences, Application Programming Interfaces (APIs), data pipelines, model integration layers, and cloud infrastructure that connect research prototypes to real product environments.
  3. Define and implement AI-driven processes that accelerate how research moves into products, applying Large Language Models (LLMs), agentic workflows, and modern developer tooling to improve experimentation, implementation, testing, evaluation, documentation, and iterative refinement.
  4. Design and integrate agentic AI frameworks and LLM-based pipelines into research tools and engineering workflows, building systems that coordinate AI agents for complex tasks such as code generation, evaluation, debugging, and refinement.
  5. Develop tool-use and function-calling architectures that enable AI systems to interact with codebases, APIs, and data sources.

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, Python, Rust, C++, C#, C, Java, JavaScript

Nice to have

  • Master's Degree or Ph.D. in Computer Science, Operations Research, Applied Mathematics, or a related field AND 10+ years of technical engineering experience OR Bachelor's Degree AND 12+ years of equivalent experience
  • Experience applying Artificial Intelligence (AI) or Machine Learning (ML) to real systems, products, or engineering workflows, with working knowledge of how modern models are trained, adapted, evaluated, optimized, or deployed, and the curiosity and ability to learn quickly across unfamiliar domains.
  • Familiarity with several of the following: agentic AI frameworks, tool use and function calling, prompt engineering, retrieval-augmented generation, evaluation frameworks, model fine-tuning or post-training, observability, and AI-assisted code generation.
  • Demonstrated engineering fundamentals and broad technical range, with experience designing, building, and shipping software systems, taking ambiguous, early-stage ideas from prototypes through production across layers such as user-facing experiences, Application Programming Interfaces (APIs) and services, data workflows, cloud systems, and model integration.
  • Experience providing technical leadership across cross-functional efforts—defining direction, mentoring engineers, and driving execution on complex projects—and building systems that balance speed of experimentation with production requirements such as reliability, security, privacy, and maintainability.
  • Proficiency with a major cloud and AI platform stack; experience with Azure, Azure AI services, and Copilot-based tools.
  • Contributions to research papers, patents, or open-source projects at the intersection of engineering and research.

What the JD emphasized

  • technical leadership
  • individual contributor
  • AI Agentic Core Team
  • accelerate the path from research to product
  • AI-driven systems, workflows, and platforms
  • researchers and product teams
  • exploration to real-world impact
  • engineers, researchers, and product teams
  • early-stage prototypes through production-ready tools, services, and experiences
  • design, build, evaluate, and ship software
  • product experiences and services
  • data, cloud infrastructure, and AI-enabled workflows
  • applying AI to real engineering and product problems
  • Deep expertise in how models work is valuable but not required
  • Research-to-Product Engineering & Technical Leadership
  • production-quality tools, services, and product capabilities
  • end-to-end engineering lifecycle
  • prototype to scalable, maintainable software
  • product experiences, Application Programming Interfaces (APIs), data pipelines, model integration layers, and cloud infrastructure
  • research prototypes to real product environments
  • Partner with MSR researchers and Microsoft product teams
  • align engineering efforts with research priorities
  • reusable components and shared architectures
  • technical direction, mentor engineers, and guide architecture and design decisions
  • best practices for code quality, security, privacy, testing, observability, and reproducibility
  • shape the team's engineering culture, long-term strategy, and adoption of modern AI-driven engineering practices
  • AI Agentic Systems & Accelerated Research Pipeline
  • AI-driven processes that accelerate how research moves into products
  • Large Language Models (LLMs), agentic workflows, and modern developer tooling
  • experimentation, implementation, testing, evaluation, documentation, and iterative refinement
  • agentic AI frameworks and LLM-based pipelines
  • research tools and engineering workflows
  • coordinate AI agents for complex tasks
  • code generation, evaluation, debugging, and refinement
  • tool-use and function-calling architectures
  • interact with codebases, APIs, and data sources
  • prompt design, retrieval-augmented generation, and evaluation frameworks
  • model experimentation and adaptation, such as fine-tuning
  • Required Qualifications
  • 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, Python, Rust, C++, C#, C, Java, JavaScript
  • Preferred Qualifications
  • Master's Degree or Ph.D. in Computer Science, Operations Research, Applied Mathematics, or a related field AND 10+ years of technical engineering experience
  • Bachelor's Degree AND 12+ years of equivalent experience
  • applying Artificial Intelligence (AI) or Machine Learning (ML) to real systems, products, or engineering workflows
  • working knowledge of how modern models are trained, adapted, evaluated, optimized, or deployed
  • curiosity and ability to learn quickly across unfamiliar domains
  • agentic AI frameworks, tool use and function calling, prompt engineering, retrieval-augmented generation, evaluation frameworks, model fine-tuning or post-training, observability, and AI-assisted code generation
  • engineering fundamentals and broad technical range
  • designing, building, and shipping software systems
  • ambiguous, early-stage ideas from prototypes through production
  • user-facing experiences, Application Programming Interfaces (APIs) and services, data workflows, cloud systems, and model integration
  • technical leadership across cross-functional efforts
  • defining direction, mentoring engineers, and driving execution on complex projects
  • systems that balance speed of experimentation with production requirements such as reliability, security, privacy, and maintainability
  • Proficiency with a major cloud and AI platform stack
  • experience with Azure, Azure AI services, and Copilot-based tools
  • Contributions to research papers, patents, or open-source projects at the intersection of engineering and research

Other signals

  • building AI-driven systems
  • accelerate the path from research to product
  • design, develop, and ship systems that transition Microsoft Research (MSR) concepts into production-quality tools, services, and product capabilities
  • Define and implement AI-driven processes that accelerate how research moves into products
  • Design and integrate agentic AI frameworks and LLM-based pipelines into research tools and engineering workflows