Principal Software Engineer

Microsoft 365 and Copilot run on Substrate & Work IQ (SWIQ) — the trusted data and intelligence platform behind Microsoft 365. SWIQ is the secure foundation that lets AI understand the context of real work — people, documents, meetings, and business processes — and act on it safely, while keeping enterprise permissions, security, and compliance intact. It is what makes Microsoft 365 Copilot and the coming wave of AI agents useful and trustworthy for hundreds of millions of people. Inside SWIQ sits OIC — Observability & Intelligent Cloud — the team responsible for keeping one of the world’s largest and most complex SaaS estates, Microsoft Exchange and Copilot, healthy in real time. OIC watches the signals these services emit, detects trouble before customers feel it, diagnoses what is actually wrong, and drives the fastest possible path to mitigation.

We are looking for a Principal Software Engineer to join our team. As a delivering Principal, you will personally prototype the next generation of customer experiences across the full incident lifecycle — Signals, Detection, Diagnosis, Mitigation, and the Incident experience — putting agentic AI to work on real operational problems and getting the earliest hands on Microsoft’s newest AI technologies.

You will operate at the center of an established, international engineering team of more than 100 people, setting technical direction not by authority but by influence — earning the trust of peer Principal leaders and pulling the organization forward through what you ship. This is a big-data role at its core: you will mine vast existing telemetry systems and invent entirely new telemetry designed for AIOps rather than inherited from human dashboards. You will apply classic AI/ML to separate genuinely concerning signals from an ocean of noise, then invent the systems that discover context around those signals — explaining not just that something is aberrant, but why, and what it means. And you will lead the harder cultural change: getting a large, careful organization comfortable letting LLM-based systems lead the diagnosis and mitigation of live incidents, methodically reducing the role of the human in traditionally human-in-the-loop practices until the machine is doing the work and the human is verifying, not driving.

We are looking for an inventor — someone with a track record of seeing where the field is going before it arrives, and preparing their organization to be there first. The next phase of operations is not about adding more tools; it is about building a system that thinks. Imagine an incident that detects itself, explains itself, mitigates itself, and files its own retrospective — while an on-call engineer sleeps. Imagine telemetry that is designed to be reasoned over by models, not just read by people. Imagine an operations practice where “human-in-the-loop” gradually becomes “human-on-the-loop,” and then simply “human-informed.” That is the space you will define. If you are energized by ambiguous, high-stakes, first-of-their-kind problems — by proving what is possible at planetary scale and then making it the new normal — this is a rare seat at the frontier of agentic AI doing real work.

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.