Jun 2016 – Jan 2026 · 9 yrs 7 mos · Mountain View, CA

• 2021 - 2026: AlloyDB for PostgreSQL
• Founding engineer; built and launched AlloyDB, a fully managed PostgreSQL-compatible database service for the most demanding enterprise database workloads. AlloyDB combines the best of Google with open-source PostgreSQL for superior performance, scale, and availability.
• More details: https://cloud.google.com/alloydb
• 2017-2021: Napa
• Founding engineer; built and launched Google's next-generation Data Warehouse, which is in critical path of majority of Google's revenue.
• More details: https://research.google/pubs/napa-powering-scalable-data-warehousing-with-robust-query-performance-at-google/
• Main projects
• Designed and built a distributed scheduling and execution framework for large-scale data processing (~XX PB/day). The system included a multi-homed task-master and a sophisticated scheduler to prioritize and manage workloads efficiently
• TL for Napa DML (Data Manipulation Language) which added support for update/deletes and atomic delete+inserts to Napa. Scaled it to be 5x faster than predecessor (Mesa) while having a better UX.
• TL for security and privacy; drove efforts around encryption, wipeout, privacy certifications, etc.
• 2016-2017: Core Search Indexing
• Worked on a system for data indexing, storage and processing, which powers general indexing needs at Google. This system is responsible for building the search index for most of Google’s major products: Search, YouTube, Shopping, Play Store, Maps, Gmail, Photos, etc.
• Main project:
• Built online-indexing frameworks to accelerate experimentation for search quality teams. Reduced iteration time from a week to 20mins for first diffs, and allowed researchers to do instant debugging in search stack for both code and data changes. This project required broad code changes and performance improvements (retuning, multi-tier caching, etc.) to restore performance during experiments.

Jan 2016 – May 2016 · 4 mos · Greater Pittsburgh Area

• TA for the graduate level course 18-842: Distributed Systems offered by the department of Electrical and Computer Engineering.

Jan 2015 – Dec 2015 · 11 mos · Greater Pittsburgh Area

• TA for the course 15-619: Cloud Computing offered by Computer Science Department.
• Over two semesters, I worked on development of 7 projects for the course, including 3 brand new ones. These projects included writing a load balancer with custom scheduling strategies, implementing an intelligent caching server, stream processing with Kafka and Samza, and using AWS APIs for autoscaling and load balancing
• I also delivered recitations, answered student queries on Piazza, and held office hours for over 160 students in Spring'15 and over 250 students in Fall'15.

Sep 2015 – Dec 2015 · 3 mos · San Jose

• Project was about building cloud-scale microservices. Team consisted of five CMU students and two Ericsson researchers

May 2015 – Aug 2015 · 3 mos · Seattle, WA

• Interned in the team building Amazon's next-generation Inventory Platform, a large scale distributed system, that will become the authoritative source of all Inventory across Amazon.

Jul 2013 – Jul 2014 · 1 yr · Bangalore

• Applications on the Salesforce.com platform for the global sales and marketing teams.
• Data integration between different systems (Salesforce.com, data warehouse, etc).

May 2012 – Jul 2012 · 2 mos · New Delhi Area, India

• Interned in Directorate of Management Information Systems and Technology (DMIST). I worked on a project aimed at predicting a blogger's age group and gender from the analysis of their blog. I achieved a peak accuracy of over 80%, and presented a paper describing the work at HIS 2012.