Amit Shankar
SWE @ Microsoft Teams
- Role
- Software Engineer 2 at Microsoft
- Location
- Seattle, WA, US
- LinkedIn followers
- 500 followers
About Amit Shankar
A full stack developer who is extremely passionate about building AI-powered performant and delightful user experiences. Working on Microsoft Teams to enable new capabilities in the real-time communication pipeline.
Experience
Software Engineer 2
Aug 2019 — Present · Bellevue, WA, US
Nov 2020 - Present: SWE on the Microsoft Teams Client Data Layer and Sync team, owning the real-time communications pipeline of the web and desktop apps, used by 250+ million MAU- Implemented a more efficient WebSocket-based last-mile notification delivery mechanism for real-time updates on the Teams client reducing missing messages by 40% and server-side costs by more than $2 million per month- Cut server load by >50% across key services by leading migration from polling to push-based sync for conversation folders, calendar events and other data entities; enabled org-wide adoption through architecture support and close cross-team collaboration- Improved chat switch reliability to 99+% for 14M daily chat/channel switches by designing and rolling out gap detection metrics; enabled root-cause investigations across client and server layers, resolving 140K+ daily message drop cases- Implemented a more efficient way to fetch chat and channel conversations during app start. Migrated core services to consume the new fetch pipeline, which led to an improvement in app launch times of 50%- Contributed to open source (https://github.com/microsoft/ObjectStoreProvider) and prototypes in Rust/WASM during hackathons2019-20: SWE in the Microsoft Azure Commerce organization (Cloud + AI)- Designed and developed a new auditing service to replace legacy systems and meet government and secret cloud requirements such as JEDI. Built end-to-end pipelines for the audit service and other pre-existing microservices to fully automate deployment, provisioning and monitoring of the services and their dependencies to scale out to public, government and secret clouds. The new service helped meet strict compliance requirements and reduced API latency by 90% compared to the old endpoint- Created an Azure-based QnA bot to enable self-serve help for commonly asked questions by customer success agents. This helped reduce incident volumes by 50%.
Education
University of Arizona
Bachelor of Science (BS), Computer Science
2015 — 2019
Find verified contacts for anyone on LinkedIn
Unifers gives sales teams verified emails and direct dials, enriched profiles, and outreach that lands in the inbox.
Free plan included · No credit card required
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.