Mar 2
Loading video player...
Get all the code from GitHub: https://github.com/SIDDHANTSAXENA2004/15-days-python-master-course/tree/main/day%2014 In this one-shot tutorial, we build a complete, full-stack Generative AI chatbot from scratch! This isn't just a simple demo. You'll learn how to build a powerful RAG (Retrieval-Augmented Generation) application with a separate frontend and backend, and give it PERMANENT memory using a SQL database. We'll build a FastAPI backend to handle all our AI logic, using LangChain and Google's Gemini model to answer questions based on our own custom documents. Then, we'll build a beautiful Streamlit frontend for our chat interface, complete with a user "login" system. The best part? We'll connect it all to a PostgreSQL database to store user information and save every single conversation. This gives our chatbot true, long-term memory for every user! 🔥 By the end of this video, you'll have a complete, portfolio-ready project. --- ► TECH STACK USED: --- • Backend: FastAPI, Uvicorn • Frontend: Streamlit, Requests • AI / RAG: LangChain, Google Gemini (gemini-2.5-flash) • Embeddings: HuggingFace (sentence-transformers/all-MiniLM-L6-v2) • Vector Store: FAISS • Database: PostgreSQL (with psycopg2) • Environment: Python-dotenv --- ► WHAT YOU'LL LEARN: --- • How to structure a full-stack AI project (backend/frontend separation). • How to create a PostgreSQL schema for users and chat history. • How to ingest custom PDFs and .txt files into a FAISS vector store. • How to build a complete RAG chain with LangChain and Gemini. • How to make your RAG chain "stateful" by passing in chat history. • How to build a REST API with FastAPI for login, history, and queries. • How to build a multi-page Streamlit app with a simple login. • How to connect a Streamlit frontend to a FastAPI backend. • How to give your chatbot true, persistent memory with SQL. --- 0:00 - Project Introduction & Final Demo 5:10 - Technology Stack & Project Setup 16:21 - Database Setup (Creating PostgreSQL Tables) 29:26 - Data Ingestion (Creating FAISS Vector Index) 37:35 - Backend (FastAPI): Building the RAG chain and API endpoints (Login, History, Query). 1:12:58 - Frontend (Streamlit): Building the UI, managing session state, and calling the API. 1:39:07 - Final Full-Stack Demo & Database Verification Thanks for building with me! If this video was helpful, please like, subscribe, and ask any questions in the comments below! #Python #RAG #Chatbot #FastAPI #Streamlit #PostgreSQL #LangChain #GoogleGemini #MachineLearning #AI #FullStack #VectorDatabase #FAISS #ProjectTutorial