SH
SHELL
AI OS PORTAL
Back to hands-on classes
Coming upIntermediateOnline

Build end to end RAG with LangChain or AWS Knowledge Bases

Directly connect your data to LLMs. Build robust Retrieval Augmented Generation systems using modern stacks.

Schedule

July 11 - July 18, 2026

Duration

2 weeks • 3hrs/wk

Project

Hands-on capstone

Detailed Curriculum

4 practical sections built around live exercises.

01

RAG architecture and embeddings

Learn how retrieval grounds model responses in your own documents.

Topics covered

  • RAG system components
  • Embeddings and similarity search
  • Chunking strategies
  • Metadata and filtering

Hands-on lab

Index a small document collection and retrieve relevant chunks for a user question.

02

Document ingestion pipelines

Prepare messy source material for dependable retrieval.

Topics covered

  • PDF and web document loading
  • Chunk quality
  • Vector database choices
  • Refresh and re-indexing patterns

Hands-on lab

Build a repeatable ingestion flow from documents to searchable vectors.

03

Generation, evaluation, and safety

Turn retrieval into grounded answers with measurement and guardrails.

Topics covered

  • Prompting with retrieved context
  • Source citation
  • RAGAS-style evaluation
  • Failure cases and fallback behavior

Hands-on lab

Evaluate answer quality and improve the retrieval prompt.

04

AWS Knowledge Bases vs custom RAG

Compare managed and custom approaches so you can choose the right production path.

Topics covered

  • AWS Knowledge Bases workflow
  • LangChain architecture
  • Cost and operations tradeoffs
  • Deployment decision matrix

Hands-on lab

Design an implementation plan for a real company knowledge assistant.

What You Get Out Of It

Concrete capabilities you should leave with.

Build a complete RAG pipeline

Understand vector search and chunking tradeoffs

Evaluate RAG quality instead of guessing

Choose between managed AWS and custom LangChain stacks