I've built and evaluated production Conversational AI at Zebra Technologies, automating LLM, RAG, and agentic AI evaluation across 1,000+ test cases using DeepEval, RAGAS, and Llama 3.1.
At Ninjacart, I contributed to Intelli-Agent for Credit, Risk, and Screening, combining XGBoost, FinGPT, OpenAI LLMs, RAG, LangChain, LangGraph, prompt engineering, and fine-tuning to support context-aware decisions and financial report generation. I also developed an XGBoost loan-eligibility model that outperformed Logistic Regression.
Alongside AI work, I've built enterprise workflow automation with Java, Spring Boot, and REST APIs, including authentication, user management, service configuration, and Google Sheets integrations.
I've also created an LLM-powered Text-to-SQL application and sentiment analysis models, and published IEEE research on customer churn prediction using AI and big data analytics.
