Built a Retrieval-Augmented Generation (RAG)-inspired chatbot as part of my portfolio, featuring a custom-built knowledge base for semantic information retrieval and NLP experimentation. Implemented text embedding generation using the Google Gemini Embeddings API with manually structured and indexed document chunks, and developed a cosine similarity-based semantic search pipeline with type-aware matching to retrieve the most relevant context. Integrated the retrieved data into a prompt engineering workflow with Google Gemini to generate accurate, context-aware responses, demonstrating practical experience in LLM integration, embeddings, and retrieval-based AI systems.