Clickhouse DB Wrapper: Led the design and development of a library to convert query requests into optimized SQL, supporting subqueries and joins. Guided multiple teams in migrating from Elasticsearch, reducing costs and improving security.
Scheduler Service: Drove scheduler optimization by fixing critical bugs and implementing scalable solutions, ensuring millisecond execution for 1M+ jobs. Helped to design a new architecture to enhance efficiency, reducing Dynamo RCU/WCU consumption.
Spearheaded the migration of 5TB of scheduled job data from the old to the new service, optimizing performance. Improved searchability by syncing live data across Elasticsearch and DynamoDB using Dynamo Steams.
PDF Generation Service: Architected and developed a high-performance PDF generation system using Typst CLI, AWS Lambda, and GraalVM-cutting processing time to seconds. Build scalable APIs for template creation, S3 storage, and retrieval.
Universal Query Language: Engineered UQL for unified querying across MongoDB, DynamoDB, and Elasticsearch, eliminating the need for developers to learn each database and write custom repositories. Managed query and result serialization/ deserialization, request authentication, and query redirection using interceptors.
MongoDB REST Service: Designed a reactive service for MongoDB, centralizing database connections and replacing native implementation in microservices. Contributed to the metadata service for managing type definitions and index metadata.
One Click Env Setup: Created service to automate cluster setup, including core services and gateways, deploying 700 microservices with configurations and routes, streamlining setup. Provided critical debugging and infra support for deployment.
Kafka Support: Led Kafka optimization strategies, mentoring teams on best practices and debugging critical issues. Enhanced Kafka library functionalities to improve performance and scalability across multiple projects.