This work is part of a RAG-based AI project. I used LangChain, Hugging Face Embeddings, and ChromaDB to build a vector store. The documents are converted into numerical embeddings and stored in ChromaDB along with their metadata. Then, when a user asks a question, the system performs similarity search to find the most relevant documents based on semantic meaning. I also converted the vector store into a retriever, which can be integrated with an LLM to build a complete document question-answering system.