Course Outline for Generative AI with LangChain

Goal:

The goal of this Generative AI with LangChain course is to provide learners with a strong understanding of generative AI concepts, LangChain components, retrieval-augmented generation, chatbot development, agents, tools, and end-to-end GenAI product building.

Audience:

This course is designed for learners, developers, data professionals, and aspiring AI engineers who want to gain practical knowledge of generative AI and build real-world applications using LangChain.

Pre-requisites:

Basic Python knowledge is recommended. A 2-hour Python refresher before Week 1 may be included for students who need additional preparation.

Duration:

12 weeks / 48 hours

Course Content

Module 01: GenAI Foundations and Introduction

Duration: Week 1–2 / 8 hours

• GenAI roadmap: big picture, career paths, and industry use cases
• Introduction to LangChain
• Why LangChain is used
• Overview of the LangChain ecosystem
• Tools and platforms: OpenAI, Hugging Face, Google, and Meta
• End-to-end GenAI product story
• Understanding how real GenAI products are built

Assignment: Build a basic “hello world” LLM application to understand how prompts, models, and outputs work.

Module 02: LangChain Core Components

Duration: Week 3–5 / 12 hours

• LangChain components
• Full architecture overview of LangChain
• LangChain models
• In-depth model tutorial with code demo
• Prompts in LangChain
• Prompt templates
• Few-shot prompting
• Dynamic prompts
• Structured output in LangChain
• Output parsers in LangChain
• Chains in LangChain
• LCEL and sequential chains
• Runnables in LangChain: Part 1 and Part 2

Assignment: Build a simple Q&A chain from scratch to reinforce Runnables and Chains together.

Module 03: Data Pipeline and RAG

Duration: Week 6–8 / 12 hours

• Document loaders in LangChain
• Loading data from PDFs, websites, CSV files, and databases
• Text splitters in LangChain
• Chunking strategies
• Vector stores in LangChain
• Embeddings
• Using Pinecone, FAISS, and Chroma
• Retrievers in LangChain
• Retrieval strategies
• RAG explained
• What RAG is and how it works
• RAG architecture deep-dive

Assignment: Prepare an “Embedding Models Comparison” presentation covering OpenAI, Hugging Face, Google, and other commonly used embedding models.

Module 04: Agents, Tools, and Chatbot Development

Duration: Week 9–10 / 8 hours

• Tools in LangChain
• Built-in tools
• Custom tools
• Tool calling in LangChain
• Function calling
• Tool-use patterns
• Building a RAG chatbot system
• Hands-on full chatbot implementation

Assignment: Build a small chatbot or agent workflow based on a student-selected interest area.

Module 05: Capstone Project: Medical Chatbot

Duration: Week 11–12 / 8 hours

• End-to-end project architecture planning and design
• LLM integration
• OpenAI and open-source model selection
• Pinecone vector database setup
• Data ingestion
• LangChain pipeline development
• Retrievers, chains, and tool calling
• Symptom analysis system
• Context-aware response system
• Deployment using Flask or FastAPI backend
• Scalable hosting overview

Assignment: Complete an “Evaluation and Testing” exercise to evaluate chatbot quality using RAGAS or similar frameworks.

Final Project

• Students will build and deploy a Medical Chatbot using LangChain, RAG, vector databases, and LLM integration.
• Code review and project demo will be conducted as part of course completion.
• A certificate of completion will be issued after successfully completing all required assignments and projects.