At TechBulls SoftTech, I developed Retrieval-Augmented Generation applications using open-source Small Language Models such as Gemma for domain-specific tasks.
I designed end-to-end LangChain pipelines for document ingestion, vector storage, and query processing. I evaluated answer relevance, faithfulness, and context precision with RAGAS, and improved retrieval accuracy through chunking strategies and embedding techniques.
For my QA ChatBot project, I built a RAG-based tool for company policy and guideline questions, using ChromaDB for document retrieval and RAGAS to evaluate response quality. I also developed a Streamlit interface for real-time queries.
I developed Blog Writing Agent to plan, research, and generate structured blog content, and LegalGuard to analyze legal documents for risks, inconsistencies, and missing clauses. I’m interested in building intelligent applications with generative AI.
