Oct 2023 – Jul 2025 · 1 yr 9 mos · Bengaluru · On-site

• Was Part of the Events Platform team, managing 16+ ProductionKafka clusters spanning 200+ brokers, supporting 1.2M+ messages/sec and 1 GBps throughput.
• Led GCP → Tencent Cloud migration, reducing infra costs while ensuring high availability, scalability, and production-grade performance.
• Designed and built Kafka Governance tool as the single source of truth for Kafka inventory, enabling declarative workflows for Kafka entities, ACLs(POC), and cross-cluster synchronisation.
• Carried out key POCs: running Kafka on Kubernetes (Strimzi) with upgrades/monitoring/secure access, and optimising MirrorMaker 2 replication latency to under 70ms, close to the 65ms RTT between GCP Taiwan and Tencent Jakarta.

Aug 2023 – Oct 2023 · 2 mos

• Developed and implemented a high-performance Apache Flink job for real-time data aggregation, boosting inline analytics for bot interactions. Collaborated cross-functionally to define metrics, optimize latency, and ensure fault tolerance.
• Built a robust reverse proxy gateway microservice with advanced capabilities for API request orchestration, including security, monitoring, and routing.
• Revamped the BotAnalytics product architecture, migrating from Kafka/SQL to Apache Pinot and Presto, scaling the system's capacity from 100K to 1M messages/day.
• Developed a data export feature enabling analytics reports to be generated via Pinot (aggregation) and Presto (multi-filter queries), exported to Excel, and made accessible via a GCP-hosted dashboard.
• Created the Conversation Path feature using a tree-level data structure to help users visualize and analyze interaction flows.
• Managed and improved over 11 microservices/modules, playing a key role in integrating bot analytics with Apache Presto and enhancing overall system observability.

Aug 2022 – Jul 2023 · 11 mos

• Working with Bot Analytics Team

Jan 2022 – Aug 2022 · 7 mos

• Designed, developed, and tested a batch processing prototype within a microservices architecture, significantly improving Cassandra DB performance. Integrated a Kafka Sink Connector to reduce throughput lag from 100,000 events to under a few hundred.
• Built a service for automated file ingestion from an SFTP server, coordinating with multiple microservices to parse file content and generate actionable tickets.
• Maintained and enhanced the Elastic Indexer microservice, enabling real-time indexing of Kafka stream data into an Elasticsearch cluster for efficient search and analytics.

Sep 2021 – Nov 2021 · 2 mos · Bangalore Urban, Karnataka, India