Jul 2025 – Present · 11 mos

• Leading engineering for Copilot growth across Word, Excel, and PowerPoint mobile apps at Microsoft.
• Drive roadmap, architecture reviews, and A/B experimentation for Copilot growth surfaces
• Ship PLG-driven onboarding and retention experiments improving user engagement
• Own customer engagement funnels across M365 mobile experiences
• Mentor engineers, lead hiring, and build scalable onboarding for the Noida site

Aug 2024 – Jul 2025 · 11 mos

• Led engineering for Web Defense — Microsoft's global web protection service securing billions of users from phishing, tech-scam, and other web threats.
• Scaled the platform to protect 2.5B+ monthly devices and serve 15B+ daily lookups
• Owned architecture for detection pipelines, threat-signal ingestion, and mitigation flows
• Modernized infrastructure for resiliency, availability, and global failover readiness
• Reduced false positives via enhanced ML/heuristic detection logic
• Built LLM-powered triage tools that meaningfully cut investigation time and boosted engineering efficiency
• Recognized with Microsoft Excellence Awards for Leadership & Delivery

Jan 2022 – Aug 2024 · 2 yrs 7 mos

• Engineering Manager on Microsoft's Web Defense team — leading a team delivering core capabilities for the global web protection service.
• Led a team of senior engineers shipping detection-pipeline features and reliability improvements across the Web Defense platform
• Drove false-positive reduction by partnering with data science to enhance ML/heuristic detection logic
• Owned operational quality: on-call rigor, alerting, post-incident reviews, and reduced incident time-to-mitigate
• Partnered with PM, data science, and adjacent platform teams on cross-org launches and threat-response programs
• Hired, mentored, and grew the engineering team across geos; ran career development for ICs

Mar 2020 – Jan 2022 · 1 yr 10 mos

• IC engineer on Microsoft's Web Defense team — built backend services on Azure powering real-time URL/threat lookups for the global web protection platform.
• Built and shipped high-throughput services and REST APIs serving real-time lookup traffic
• Owned components in the threat-signal processing pipeline; tuned performance, memory footprint, and dependency cost
• Implemented telemetry, dashboards, and alerting that improved on-call detectability and response times
• Integrated ML model outputs into production detection paths in close partnership with data scientists
• Drove incident triage, runbooks, and post-incident reviews; raised the on-call quality bar