lokesh hongure

At Sumago Infotech, I applied Python, machine learning, and NLP techniques to build and evaluate models, and built an AI-powered Financial Analyser chatbot as my internship's key deliverable.

For the chatbot, I used LangChain and FAISS to process financial PDFs, create embeddings, and retrieve relevant content for context-aware answers. I also designed prompt templates and a Streamlit interface for interactive document Q&A.

On Ventryx AI, I developed a company knowledge assistant that answers natural-language questions about uploaded PDFs and grounds responses with document and page-level citations. I tested it across document types and query phrasings, and I'm focused on building practical, testable LLM-powered assistants.