LangChain's Role in the LLM App Stack
By the end of this module you can say what LangChain actually adds to an LLM application, and judge when calling the provider SDK directly is the simpler and better choice.
LangChain for LLM Applications and RAG
A traceable RAG assistant with source citations, structured answers, retrieval evaluation cases, and debugging notes.
Models, Messages, Prompts, and Provider Configuration
By the end of this module you can configure models, messages and prompts across providers, and swap one provider for another without rewriting your application.
Structured Output and Parsers
By the end of this module you can get structured output through parsers, and handle the case where parsing fails rather than assuming it will not.
Document Loading, Splitting, and Embeddings
By the end of this module you can load and split documents so chunks stay meaningful, and choose a chunk size from what the content is rather than from a default.
Vector Stores and Retrievers
By the end of this module you can put chunks into a vector store and configure a retriever, including the metadata filters that similarity search alone cannot do.
Retrieval Chains with Grounded Responses
By the end of this module you can build a retrieval chain that answers only from what it retrieved, and shows the sources it used.
Tools and Simple Agents
By the end of this module you can give a chain tools and build a simple agent, keeping each tool narrow enough that a wrong call is cheap.
Memory and Session State Boundaries
By the end of this module you can manage memory and session boundaries, deciding what persists between turns and what must not leak between users.
Streaming and Callbacks
By the end of this module you can stream responses and use callbacks to observe a chain as it runs, rather than only seeing what it returned.
Tracing and Debugging with LangSmith
By the end of this module you can trace a chain in LangSmith and use the trace to find where a wrong answer was actually introduced.
Testing and Evaluating LangChain Components
By the end of this module you can test LangChain components in isolation with fake models, so your suite runs fast, free and deterministically.
Capstone: Traceable RAG Assistant
In this capstone you will build a traceable RAG assistant with grounded citations, a working trace for every request, and tests that run without calling a paid API.
First-release boundary
This is a text-first course shell. Leonardo/image creation, image QC, dependency-heavy runtime QA, external review, and learner pilot evidence remain separate later quality steps.