Part of AdServing infrastructure team responsible for serving Ads on several Google-owned properties like Google.com, Youtube with tight latency budget and large QPS.

TL in Data Infrastructure and Analytics group (formerly Ads Data Infrastructure). Founder and lead of a new highly scalable streaming infrastructure for generating very low-latency statistics (involving log-joins, DB joins, and low latency aggregation) for business critical use-cases in and outside Google.

Member of Evenflow team that powered critical Ads metrics used for Ads reporting, billing, and internal use cases at Google.

designed a bandwidth-aware reactive system for scheduling copies of big data over WAN to drastically improve reliability while radically reducing WAN bandwidth utilization.

Worked on Paxos. Created a new storage layer optimized for a large Ads use case to drive significant resource savings.