Developed a product to logically identify effectiveness of a marketing campaign based on various data sources using SQL and Java. Optimized query execution plans by analyzing query patterns which reduced the query runtime by 80%. Led a team of 2 professionals in developing a comprehensive product roadmap, aligning it with the company's strategic goals. Served as the primary liaison and point of contact for stakeholders during the analysis of marketing experiments. Ensured timely and accurate reporting of experiment analysis results to stakeholders, facilitating informed decision-making and driving data-driven marketing strategies.

Designed and implemented real-time machine learning platform in both streaming and batch processing modes using AWS, Java, Flink, and Beam. Improved ingestion pipeline algorithm efficiency by 80% in AWS. This improvement changed the service from Batch to Lambda which reduced the cost by 75%. Designed and implemented Reference Data layer in the platform for Semantic Interoperability. This involved data ingestion, transformation, aggregation, and querying from multiple sources in AWS Athena with AWS Glue Catalog. Designed SNS, SQS, Glue, and Sagemaker I/O in Beam for convenient integration inside the streaming pipeline deployed in AWS EMR. Investigated Apache Flink on AWS Batch as a mini-cluster and on EMR as a distributed cluster on its cost, maintenance, operation, and performance.

Developed modularized framework on top of MithrilJS (Virtual DOM) to deliver responsive web applications. The framework optimized the clinical workflow components performance and interoperability using ES6 (ES2015), parallelly reducing the technical debt on the codebase. Enhanced the framework by creating reusable modules that reduced the development timeline of clinical components.