RAG Project: Build an AI Onboarding Chatbot with Streamlit, LangChain, and ChromaDB

Video Description

In this comprehensive video, you'll learn how to build a fully functional AI-powered onboarding chatbot tailored for new employees at a fictional company (Umbrella Corporation). The tutorial walks through each stage of development, from managing environment variables and parsing company policy PDFs, to implementing retrieval-augmented generation (RAG) pipelines using LangChain, integrating vector search with Chroma, building a rich frontend with Streamlit, and instrumenting your app for traceability with LangSmith. Ideal for AI engineers, data scientists, and Python developers, the video provides hands-on implementation details, best practices, and debugging insights, making it a perfect portfolio project or a foundation for enterprise onboarding solutions. ## Topics - Working with OpenAI and Groq API keys - Understanding and using LangSmith for tracing/debugging - Using Faker for generating sample employee data - Parsing and extracting text from PDFs using PyPDF - Handling structured and unstructured PDF data - Loading, vectorizing, and splitting documents using LangChain and text splitters - Creating and managing Chroma vector stores for document retrieval - Building RAG (Retrieval Augmented Generation) applications ## Links and code *Important*: This video is a module from the AI Engineering Bootcamp. Sign up here: - 🚀 Complete AI Engineer Bootcamp: https://aibootcamp.dev 👉 Code (starting point): https://github.com/alejandro-ao/client-onboarding-rag-demo 👉 Code (solution): https://github.com/alejandro-ao/client-onboarding-rag-demo/tree/solution - ❤️ Buy me a coffee... or a beer (thanks): https://link.alejandro-ao.com/l83gNq - 💬 Join the Discord Help Server: https://link.alejandro-ao.com/HrFKZn - ✉️ Get the news from the channel and AI Engineering: https://link.alejandro-ao.com/AIIguB ## Timestamps 0:00:00 - Intro 0:02:39 - Project setup 0:07:31 - Data Service 0:18:10 - Assistant Class 0:34:56 - GUI With Streamlit 0:44:35 - Run the Chatbot 0:51:22 - Tracing 0:56:56 - Conclusion