Building distributed systems that stay up at scale.

Staff Software Engineer · Databricks

Principal Software Engineer · Swiggy

• Led a revamp of the catalog ingestion pipeline: simpler flows, better observability, over 60% fewer frequently occurring issues and roughly 50% faster new-feature development. Proposed and drove the move from MySQL to DynamoDB for elasticity.
• Ran initiatives across Pricing and Promotions on scalability, operational load and resilience (circuit breakers, rate limiters, caching strategy, alerting and governance). S1 incidents dropped to zero over six months, with a ~60% reduction across S1/S2/S3.
• Added cron management to the in-house IaC framework on AWS Batch and onboarded 50+ jobs in six months, replacing crons modeled as web services or run as scripts on standalone EC2 instances.
• Benchmarked OpenJDK, Corretto and Oracle Java and led an organization-wide migration of 300+ services to Corretto, standardizing on Docker images in AWS ECR and removing the dependency on public Docker Hub.

Staff Engineer · Roposo (acquired by Glance, InMobi Group)

• Scaled backend infrastructure for millions of daily active users of a social network.
• Content Delivery Framework: a pluggable framework for testing new ranking algorithms, with A/B experimentation, data tracking and ~100ms latency targets.
• Real-Time Recommendation Engine: end-to-end architecture (HLD and LLD) for recommendations within ~60 seconds of user interaction, which lifted time spent on feed by about 20%.

Senior Software Engineer · GUAVUS

• Developed a user segmentation engine that required building real-time streaming pipelines with an in-house rule engine over terabytes of data, combining static user profiles with dynamic rule-based segmentation on browsing behavior, and choosing the database for storing aggregates by benchmarking HBase against Cassandra.

Software Engineer · Belzabar

• Built a medical workflow management system on Alfresco: JSF interfaces, web services, Lucene based search and ACL.