LangChain Learning App

An interactive CLI app to learn core LangChain concepts through hands-on examples.

Lessons

#	Topic	Concepts Covered
1	LLMs & Prompt Templates	Direct LLM invocation, PromptTemplate, ChatPromptTemplate, few-shot prompting
2	Chains (LCEL)	| pipe operator, StrOutputParser, RunnableParallel, RunnablePassthrough
3	Memory & Chat History	ChatMessageHistory, RunnableWithMessageHistory, multi-session isolation
4	Agents & Tools	@tool, create_react_agent (LangGraph), custom tools
5	RAG	Embeddings, InMemoryVectorStore, RAG chain, source citations, conversational RAG
6	LangGraph	StateGraph, nodes & edges, conditional branching, message state, ReAct agent from scratch
7	Output Parsers	StrOutputParser, CommaSeparatedListOutputParser, JsonOutputParser, PydanticOutputParser, with_structured_output, manual output fixing
8	Streaming	.stream(), .astream(), LangGraph streaming, .astream_events()
9	Document Loaders	TextLoader, CSVLoader, DirectoryLoader, RecursiveCharacterTextSplitter, CharacterTextSplitter, TokenTextSplitter
10	Multi-Agent Systems	Supervisor pattern, sequential handoffs, parallel fan-out agents, shared state
11	Evaluation & Testing	LLM-as-judge, criteria scoring, pairwise A/B comparison, batch dataset evaluation
12	Caching	InMemoryCache, SQLiteCache, latency comparison, cache hit rate
13	Callbacks & Tracing	BaseCallbackHandler, token usage tracking, timing, logging, multi-callback
14	LangChain + FastAPI	REST endpoints, streaming SSE, per-session chat memory, httpx tests
15	LangChain + Databases	SQLDatabase, NL→SQL chain, read-only SQL agent with create_react_agent

Setup

1. Clone the repo

git clone https://github.com/Apusapati-9/langchain-learning.git
cd langchain-learning

2. Install dependencies

pip install -r requirements.txt

3. Add your OpenAI API key

cp .env.example .env
# Edit .env and set OPENAI_API_KEY=sk-...

Usage

# Interactive menu
python main.py

# Run a specific lesson
python main.py --lesson 1

# Run all lessons
python main.py --all

Project Structure

langchain-learning/
├── main.py                  # CLI entry point
├── requirements.txt
├── .env.example
└── lessons/
    ├── 01_llm_prompts.py    # Lesson 1: LLMs & Prompt Templates
    ├── 02_chains.py         # Lesson 2: Chains (LCEL)
    ├── 03_memory.py         # Lesson 3: Memory & Chat History
    ├── 04_agents.py         # Lesson 4: Agents & Tools
    ├── 05_rag.py            # Lesson 5: RAG
    ├── 06_langgraph.py      # Lesson 6: LangGraph
    ├── 07_output_parsers.py # Lesson 7: Output Parsers
    ├── 08_streaming.py      # Lesson 8: Streaming
    ├── 09_document_loaders.py # Lesson 9: Document Loaders
    ├── 10_multi_agent.py    # Lesson 10: Multi-Agent Systems
    ├── 11_evaluation.py     # Lesson 11: Evaluation & Testing
    ├── 12_caching.py        # Lesson 12: Caching
    ├── 13_callbacks.py      # Lesson 13: Callbacks & Tracing
    ├── 14_fastapi.py        # Lesson 14: LangChain + FastAPI
    └── 15_databases.py      # Lesson 15: LangChain + Databases
└── data/                    # Auto-generated sample files (gitignored)

Requirements

• Python 3.10+
• OpenAI API key
• Dependencies: langchain, langchain-openai, langchain-core, langchain-community, langchain-text-splitters, langgraph, python-dotenv, fastapi, uvicorn, httpx