Harshada is part of the core engineering team for Brooklin at LinkedIn, an open-source data streaming platform used to build and operate large-scale CDC and Kafka ingestion systems in production. Brooklin runs across more than 50 clusters, mirrors over 122k+ Kafka topics, and processes more than 7 trillion messages per day while supporting sub-minute SLAs for critical workloads.
Their work focuses on designing reliable, scalable ingestion architectures that support heterogeneous databases and shared, multi-tenant streaming platforms. They have hands-on experience with Apache Flink and Apache Samza, which consume CDC data produced by Brooklin for real-time analytics and machine learning use cases.
Harshada has presented at internal engineering conferences at LinkedIn and Goldman Sachs, sharing real-world lessons from building and operating large distributed systems.