Aishwarya Bazzurla

At Tata Consultancy Services, I architected and deployed an LLM-powered RAG recommendation platform for Google’s internal SkillCraft learning ecosystem, enabling semantic discovery across 10,000+ learning resources.

I engineered semantic retrieval workflows using dense vector embeddings and Cloud Spanner vector search, and designed ETL pipelines to ingest, normalize, and index learning materials. I also developed Python backend services and owned the implementation of PLX analytics dashboards with role-based authorization.

On my AI Code Reviewer project, I designed a six-node LangGraph multi-agent pipeline and a two-tier semantic cache using SHA-256 matching and pgvector. I also built an asynchronous task queue and a Redis-backed circuit breaker for LLM provider failures.

For my Plant Disease Detection project, I fine-tuned DenseNet121 on 43K+ images across 44 disease classes and 14 crops, achieving over 99% validation accuracy on the curated dataset. I deployed it through a Flask API with real-time image upload and per-disease treatment lookup.