New 
Principal Machine Learning Engineer
 Microsoft | |
 United States, Washington, Redmond | |
 Oct 14, 2025 | |
|
OverviewCore AI is at the forefront of Microsoft's mission to redefine how software is built and experienced. Weare responsible forbuilding the foundational platforms, services, programming models, and developer experiences that power the next generation of applications using Generative AI. Our work enables developers and enterprises to harness the full potential of AI to create intelligent, adaptive, and transformative software. The Observability group is focused on developing solutions to monitor, evaluate, and optimize AI agent performance. We are seeking a passionate and skilled software engineer to join the Observability platform team. This team is responsible for building the services that power Observability in Foundry.
ResponsibilitiesDesign, implement and deliver AI services to supportproductofferingsforlarge-scaleagent observability Design and build the end-to-end pipelines covering model training, data analysis, model serving and model evaluation. Design and develop scalable systems for benchmarking AI models, including pipelines for automated evaluation, metric tracking, and result visualization. Build and maintain a robust data platform to support model evaluation workflows, including ingestion, versioning, and storage of datasets and model artifacts. Demonstrate good understanding of LLM architectures, training and inference Collaborate closely with product management and partner teams to align technical direction with business goals Take end-to-end responsibility for the development lifecycle and production readiness of the services you build and drive the team's DevOps culture Engage with customers to gather feedback and resolve complex issues Understand Microsoft businesses and collaborate with stakeholders towards cohesive, end-to-end experiences for Microsoft customers Innovate ontechnical solutions, and patterns that will improve the availability, reliability, efficiency, observability, and performance of products. | |