Dagger: Developed framework for end to end run-time custom java code injection in in-house Flink based Data Streaming Platform. Built Smoke and Integration testing pipelines and added multiple other features.

Longbow: Implemented Key-value based state backend on top of Bigtable for complex Flink Streaming queries requiring up to 3 months of production Data.

Thor: Built a CLI tool for all in-house Data products. Using thor exposed git-based Data platform to customers.

Firehose: Incorporated multiple features to in-house Kafka-sink platform like Redis and Influx sinks, Templatised JSON body support to HTTP sinks and sidecar based monitoring using influx reporter.

GoMemory: Worked on a centralized real-time user profiling Database on DynamoDB and Kafka-streams to provide user details < 200ms. Built a continuous integration testing framework to ensure correctness and a lightweight web API serving user personalization data on top of it.

KShift: Added Avro and protobuf data support in Kafka to Redshift/S3 ETL tool.

Data Infrastructure: Worked on S3 + Delta-based Data warehouse. Maintained and continuously enhanced all Data infrastructures like Kafka, Yarn, Airflow etc. Established a fault-tolerant CDC pipeline using Debezium.

Worked on OCR module for medical reports. Experimented with basic ML models on patient health data.

Created a basic JWT based auth framework for user login flow.

Worked on image classifications and heat map generation modules from CCTV feed for user behavior extraction.