Sid Dahiya

Lead the engineering team that owns Microsoft Teams' feature-flighting and release-safety platform — shipping Teams to ~1 billion users a day. Driving data-driven, automated ring promotion (crash-free rate, test pass rate, performance, incidents, security coverage) in place of calendar-based gates, and setting direction for agentic release intelligence: AI promoting, pausing, and rolling back releases under human oversight, with engineers steering.

Architect of record for Microsoft Teams' feature-flighting platform, scaling release infrastructure for 1 billion daily users. Co-founded "Turbo," a fast-release pipeline that ships AI features in days instead of weeks — onboarding 8 squads for 2x release cadence — and designed test-pipeline auto-retry that restored availability to 100%, halved stuck rollouts, and lifted PR throughput 60%.

Built Office's Customer Experiences (CX) engineering team from a scattered feature crew into a high-performing group, and grew the CX customer-health metric from a dashboard into the org's de-facto quantitative signal — now part of the senior-leadership Monthly Product Review. Architected the end-to-end data flow for the org's pivot to Microsoft Copilot across Word, PowerPoint, Excel, Loop, OneDrive/SharePoint, Teams, and Outlook.

Led the Unified Telemetry team — the SDKs, compiler, and services that standardize how products across Microsoft's Experiences & Devices org collect telemetry. Established the .NET SDK (with an OpenTelemetry sink) as the de-facto replacement for the deprecated legacy SDK, adopted beyond Office, and steadied and grew the team through multiple reorgs while hiring and coaching engineers.

Moved from a Principal individual-contributor role into management to stand up and lead the newly formed Unified Telemetry team, unifying a fragmented set of platform-specific telemetry libraries into one modern, cross-platform SDK suite with the packaging and pipelines partners needed to adopt it.

Owned client telemetry, privacy, and cost-efficiency platforms used across Microsoft. Lead developer for Office's privacy-concern detection tooling — first adopted outside Office by Teams, OneDrive, and Visual Studio. Drove a legacy-data-reduction effort cutting Office's telemetry collection ~25% (saving hundreds of thousands of dollars a month), and ~2,248 CPU cores in Europe during COVID capacity pressure. Built and shipped the cross-platform telemetry SDKs — C++ core with Java and Objective-C wrappers, plus JS/TypeScript and .NET — frequently as the sole engineer.

Productionized Office's observability tooling into the Diagnostic Data Viewer — Microsoft's user-facing data-transparency experience — shipping across Win32, Mac, and iOS from investigation to shipped in five months. Rolled a privacy-compliant telemetry wrapper across ~288 Office projects, drove ITAR/GCC compliance, and improved telemetry-config core performance ~5x. Promoted to Senior Software Engineer.

Built "Office Telemetry Studio," the developer tool Office engineers relied on for the better part of a decade — the foundation later productionized into the Diagnostic Data Viewer. Fixed a customer-blocking startup crash (~3,000 → ~50 crash reports a day) and built "Dynamic Config," a runtime telemetry-sampling system on the company's experimentation platform.

Pennsylvania State University

Capital College

M.Sc. Computer Science (2013 – 2015)

B.Sc. Computer Science — Honors Program (2009 – 2013)

B.Sc. Mathematics — Honors Program (2009 – 2013)