Join Microsoft's AI Platform organization and play a pivotal role in shaping the future of AI. As a Principal Software Engineer on the Azure AI Foundry Customization team, you will contribute to the advanced platform powering Azure's innovations and Microsoft's flagship products. Our mission is to enable every AI application globally.

You'll operate within a highly technical environment, designing systems that facilitate AI learning, improvement, and scaling in production. This role involves tackling significant industry challenges to make frontier models accessible, efficient, and cost-effective for enterprises.

This position offers end-to-end ownership, from high-level APIs to granular reinforcement learning controls, and integration with Azure ML flexibility. You will deliver essential services, pipelines, ML recipes, evaluations, and integrations, leveraging Azure's robust infrastructure. Your contributions will drive advancements in cost, speed, and reliability through innovations like single-click data generation and dynamic model routing.

Showcase your achievements on platforms like Microsoft Build and Microsoft Ignite, highlighting breakthroughs in areas such as reinforcement learning for production agents, turning foundation models into production AI, and optimizing models with fine-tuning. Be at the forefront of making agentic AI economically viable for every enterprise.

Lead by example in coding, mentoring others to produce extensible, maintainable, well-tested, secure, and performant code. Continuously improve code performance, testability, maintainability, effectiveness, and cost, understanding relevant trade-offs.

Identify best practices and coding patterns, providing deep expertise in coding and validation strategies. Create and apply metrics to drive code quality and stability. Anticipate and resolve development blockers, escalating issues and proposing solutions.

Lead debugging efforts using tools, tests, logs, and telemetry. Proactively verify assumptions and resolve issues with minimal iterations. Conduct incident retrospectives to identify root causes and implement preventive measures. Drive adherence to least-access principles, privacy, and security.

Review product and test code to ensure team standards, coverage, and suitability. Provide insightful code reviews to enhance code quality and develop other engineers' skills, accelerating team development pace. Ensure code is diagnosable, reliable, testable, and maintainable before delivery.

Lead and own discussions for complex product/solution architectures and designs, exploring various options and making recommendations. Create architecture proposals and lead hypothesis testing for complex solutions. Guide less experienced team members and oversee design decisions.

Evaluate and integrate new technologies to solve problems. Ensure system architecture and designs meet performance, scalability, resiliency, disaster recovery, and COGS requirements. Uphold Microsoft's standards for security, privacy, and compliance. Mentor engineers on building upon existing work.

Create and guide others in developing a clear test strategy to ensure solution quality and prevent regressions. Ensure test plans include security testing and lead efforts to add, fix, or improve tests for speed and reliability.

Develop testable code and consider testability during design. Maintain up-to-date knowledge of testing architectures and apply them appropriately. Redesign or rearchitect untestable code sections and leverage AI tools for test automation.

A strong background with 15+ years of experience in writing production code for internet-scale services and distributed systems is preferred.

Demonstrated engineering knowledge of machine learning systems and data pipelines is essential.

Proven experience mentoring other developers, collaborating effectively with partners, and functioning as a dedicated team player is expected.

Excellent communication and presentation skills are required to effectively convey technical concepts and ideas.

Crucially, the ability to debug, read complex code, and contribute within a large codebase is a fundamental requirement for this role.