What is LangChain?

LangChain is a framework that will supercharge your LLM building experience!

Key Features of LangChain:

• Standaridise the process of building LLM applications
• Contain extensive library of components in one framework with community integrations
• Modular Development: LLM features are split into different components with an interface to pass data to one another

LangChain Components

• Chat Models: Unified interface to use different providers (e.g. OpenAI / Claude)
• Prompt Templates: Reuse and edit common prompts across the entire LLM apps.
• Output Parsers: Ensure the LLM outputs response in the correct format (e.g. JSON, Dictionary)

LangChain in Retrieval Augmented Generation

• Document Loaders: Load in external data sources with the pre-built integrations (e.g. PDF, Notion and many more...)
• Text Splitters: Split the loaded data into smaller sizes to fit to LLM context window
• Vector Databases: Embed loaded in data into vectors then store in a database for efficient information retrieval

LangGraph

• Orchasterate the different LLM components together to provide granular control over the LLM systems
• Build strong and robusts agents to automate tasks with fallback mechanisms, easy data handling and a wide variety of tools