At SplashLearn, I engineer Java services for book discovery and assignment workflows. I cut recommendation query time by about 40% by adding PostgreSQL indexes, restructuring joins, and changing the book-theme mapping.
I also built a distributed rate limiter using Redis and Lua, and developed book-management APIs and interfaces with Kafka event flows. For book assignment APIs, I migrated legacy Ruby on Rails code to Java and fixed N+1 queries, improving performance by about 30–40%.
On CortexAI, I built a Node.js and Express microservices backend with a LangGraph multi-agent system and a RAG pipeline using Qdrant. I also developed a Spring Boot microservices backend for an AI-powered fitness app, using Gemini for personalized workout recommendations.
