At Gyanvix, I prototyped an AI-driven talent-matching system to automate manual shortlisting. Its XGBoost-based scoring pipeline achieved 99% accuracy.
For TDS, I used a Redis producer-consumer queue for asynchronous demand scoring and built an event-driven orchestrator to automate the talent-matching lifecycle. I also led the TDS 2.0 architectural refactor, applying SOLID principles and mentoring teammates on clean architecture.
At Gyanvix's APS platform, I designed and documented the inventory module and built pricing logic for procurement and billing workflows. I also implemented audit logging to support traceability and reduce manual reconciliation.
On my Journal App project, I built a Spring Boot and MongoDB backend with JWT authentication, Redis Cloud caching, and scheduled weekly sentiment summaries. I also built a Java REST backend for Campus Marketplace, with JPA/Hibernate persistence, search and filtering, and media storage integration.
