Abhijeet Mishra

Developed a Java recommendation API with automated cron recomputation for a 15-day rolling score, ensuring 99.9% data freshness. Engineered a Java Kafka stream pipeline processing 20M+ events daily, using separate topologies for generic event sessionization and click/impression aggregation to drive real-time article recommendations. Constructed a real-time Game Ratings System in Java Spring Boot to capture and return like/dislike counts, increasing engagement tracking accuracy by 25%. Built Java Spring Boot APIs for the App Directory (10+ endpoints), improving response time and reducing DB load by 35% through Redis caching implementation. Designed a Debezium + Kafka CDC pipeline for email-refiner, streaming DB triggers from key tables, automated classification into blocked tables, cutting ops overhead by 40%. Delivered a suite of metadata APIs for onboarding games, reducing integration time by 60%. Implemented a game tagging system to categorize thousands of games, enabling personalized discovery and improved search by 30%. Developed an Elixir/Phoenix event collector for asynchronous ingestion into Kafka, reducing data pipeline latency by 50% and maintaining a 99.99% system uptime.

Developed automated Java Spring Boot lending scripts, processing 200+ applications/day, resulting in an average loan volume of $3M+. Automated invoice upload, disbursal, and repayment workflows using Scala Akka scheduler, boosting daily disbursal by 100%. Designed and deployed secure AWS S3 architecture for bulk document storage integrated with lender APIs, improving accuracy by 45%. Implemented AWS SQS for asynchronous processing, increasing system throughput by 30%.