AI Agents: Develop Autonomous AI Agents with LangGraph [NEW]

Learn optional Python installation steps for Windows, Mac, or Linux by following the Kingston Comm knowledge knowledge base to set up Python on your machine.

Explore what AI agents are, their motivation and advantages, and how autonomous agents combine large language models with tools, APIs, and memory to reason, search, and tailor responses.

Explore the characteristics of AI agents—autonomy, learning and adaptation, interaction, and goal-driven design—and use cases like customer service chatbots, personal assistants, AI agent data analysis, and smart homes.

Build your first ai agent in python by setting up a virtual environment, loading an OpenAI API key, and testing with a GPT 3.5 turbo chat completion.

Create your first ai agent by building a class that encapsulates the agent, manage system and user messages, and implement a prompt-driven loop of thought, action, pause, and observation.

Instantiate and run the agent, using system prompts and message stacking to drive the model to compute Earth's mass, observe results, and generate the final answer.

See how an AI agent processes a complex query by chaining prompts, tool calls, and observations to calculate the combined mass of Earth and Mars.

Automate an AI agent's workflow by looping through prompts, parsing actions with regex, and executing known actions (calculate and planet mass) via a query function to compute planetary masses.

Develop an interactive ai agent using a console prompt to set the maximum turns, input questions, and view tool-assisted planet calculations, including determining the heaviest planet and combined masses.

Build your first AI agent using a prompt driven loop of query, system prompt, observation, and tool use, then automate it with a console app and Lang Graph.

Check in with the course community by inviting feedback and questions, encouraging reviews, and promising timely answers and discussion on the discussion board to add value.

Learn line graph concepts with LangGraph, using graph-based knowledge to represent and orchestrate AI agent workflows through nodes and edges, supported by natural language processing and reasoning.

LangGraph enhances ai agents with graph-based knowledge management, improved natural language understanding, and efficient reasoning for scalable, flexible agent workflows. It supports cyclic graphs and state persistence for human-in-the-loop interactions.

Explore line graph core concepts—nodes (agents), edges, and conditional edges that drive decisions, with an entry point and end node forming a state-machine workflow.

Explore how LangGraph attaches state and data to AI agents to preserve context and enable retrieval of past information for better responses.

Build a simple ai agent with LangGraph using LangChain, manage state and messages, and assemble a runnable graph with a bot node, entry, and exit.

Create a console app using LangGraph's stream method to stream events, manage bot state, and interact with a simple AI agent, illustrating how tools extend bot capabilities.

Learn to extend a basic LangGraph agent by attaching tools, wiring a Tavileh search tool, and running the model with tools to perform web queries and continue the workflow.

Explore adding tools to an AI agent by implementing a basic tool node, using tool messages to pass tool results, and routing with conditional edges in a LangGraph graph.

Learn how to integrate built-in tools with a LangGraph AI agent, binding a travel search tool to the model, and routing tool calls via conditional edges to fetch results.

Add memory to your agent using a check pointer, threads, and a SQLite saver to save and restore context between interactions for fault-tolerant, long-running workflows.

Explore adding a human in the loop to autonomous ai agents using interrupt-before tool calls, memory checkpoints, and configurable workflows to improve reliability.

Explore building a simple ai agent with a line graph using langgraph, add tools and memory via a check pointer, and include a human-in-the-loop to inject feedback.

Build an autonomous AI agent with LangGraph to gather financials, analyze data, research competitors, and generate a performance report through iterative feedback and revisions.

Develop an autonomous ai agent that ingests a financials csv with pandas, analyzes and compares profits to competitors, and uses defined prompts to generate a financial report.

Create all nodes as functions to gather financials, analyze data, research competitors, compare performance, and write a final report within a LangGraph autonomous agent.

Create and run an AI agent by building a node-edge graph, gathering financials from a CSV with pandas, analyzing data, researching competitors, comparing performance, collecting feedback, and writing a report.

Add a Streamlit-based GUI to your AI agent using Streamlit to visualize data upload, task configuration, and iterative analysis of financial performance with competitor benchmarking.

Explore optimization techniques for AI agents, including model optimization, quantization, and fine tuning. Leverage parallel and batch processing, distributed computing, and caching to boost efficiency.

Explore building autonomous AI agents from scratch and with the LangGraph framework, learn about line graph, building blocks, and finish with a financial report writer and Streamlit GUI.

Continue learning after this course by extending your projects and implementing your own AI agent with LangGraph, and explore OpenAI documentation on GPT-3 and GPT-4 to attach large language models.