At Zepto, I lead architecture and scaling work across core order fulfillment systems, including a PostgreSQL partitioning migration that reduced active table size by 83%, cut query lookup latency by over 50%, and saved more than $150,000 annually in AWS costs.
I’ve led teams redesigning order management for multi-fulfillment processing, migrating distributed transaction orchestration from Kafka/SQS choreographies to Temporal, and isolating critical order booking flows from legacy database contention. These efforts improved recovery time, reduced manual intervention, supported 3x throughput, and reduced order booking failures by 15%.
Previously, I built backend platforms at Airmeet and LendingKart, improving event-data APIs, supporting whitelabel events, automating KYC and document verification flows, and strengthening API observability. I work across Java, Python, Go, PostgreSQL, AWS, Kafka, and distributed systems.
