LangChain and LangGraph Crash Course

Create a medium article generator that turns a topic like passive income into a medium title and article, including a subtitle and introduction for posting on Medium.

Set up a LangChain and Streamlit-based Medium article generator by configuring the OpenAI API key, installing dependencies, and launching a front-end that generates articles from a topic.

Connect to OpenAI by creating an OpenAI instance, and set the temperature to balance creativity for article titles with factual accuracy for legal summaries.

Learn to build reusable prompt templates in LangChain that accept inputs like topic and language. Use placeholders and a format to generate tailored article titles with precision across multiple languages.

Master simple chains in LangChain by building an lm chain with a prompt template. Learn how to pass topics and inputs—dictionary-based or single-variable—before comparing with sequential chains for complex tasks.

Learn to build OpenAI LLM powered apps using sequential chains that produce a medium article title and body by connecting a title chain to an article chain.

Discover how LangChain agents use GPT models with tools like Wikipedia and math through a zero-shot react agent to plan actions and answer prompts.

Build a chat with document bot that uploads pdf, word, or text files, reads and embeds them, and answers questions while preserving chat history and resets history on new file.

Apply LangChain and large language models to your documents by building a chat-based question-and-answer app with a chroma vector store, embeddings, and a retrieval chain.

Add memory to your LangChain apps by implementing a chat history with a conversational retrieval chain, saving questions and answers via session state for contextual, ongoing conversations.

Discover how to print chat history using a for loop over a session state 2d array, render it in the app, and address history clearing and features like file uploads.

Add a user file upload feature to read custom documents, supporting pdf, docs, and text files, and manage upload state with clear history, chunking, and embedding workflows.

Learn to load text, pdf, and word documents in a LangChain app by detecting file extensions and selecting the appropriate loader, with a spinner for chunking and embedding.

Enable a YouTube question answering chat by using the YouTube loader and transcript API to load data from YouTube and other sources like Wikipedia.

Explore the fundamentals of Lang Graph in this introduction, covering type annotation, elements, hollow world, multiple inputs, conditional, looping graphs, and rack agent, with ai agent exercises.

Explore type annotations and dictionaries in Python, using keys like name and age to retrieve values like Alice and 30, and learn that dictionaries do not enforce data types.

Implement a Python type dictionary to define state data with key types in a class, such as name string and age integer, boosting type safety and readability while reducing errors.

Explore union types and Python type annotations, enforce type safety with dictionaries and the type dictionary, and see how LangChain and LangGraph build robust AI agents and graphs.

Build a hello world graph in LandGraph to learn start, end, two intermediate nodes, and how data flows through connections, while setting up Python and a code editor.

Create your first graph in LangGraph by defining a typed state with a name attribute, adding an ask name node and a grid node, linking start to end, and compiling.

Visualize and run a simple hello world graph with an ascii diagram, an invoke method, and a start, ask, grid, and end nodes.

Create a multi-input LangGraph to collect a name and five transactions, analyze the total, and classify you as saver or spender for budget analysis.

Build graph in lang graph by initializing with state graph, start at asking and end at analyze, connect edges from asking to ask values and from ask values to analyze.

Create a conditional graph to route data using a router node, with an input function producing plus or minus edges, powering a tip calculator assistant that computes the final total.

Learn how to invoke the graph, supply default values, and route based on operations like plus and minus to compute final totals with tip or discount.

Extend the tip calculator example by adding an optional rounding step after tip or discount decisions, reinforcing conditional branches and graph concepts for LangChain and LangGraph.

Explore looping graphs with a conditional edge by building a password authentication flow that loops back to get password until the correct one is entered, then ends at success.

Run the graph demonstrates looping with conditional edges in LangGraph, guiding password input until success; implement the final graph and start building AI agents.

Build a simple AI agent by integrating LLMs into graphs with LangChain's ChatOpenAI, define SimpleBot, and invoke the LLM in a graph node with a while loop.