Learn Agentic AI – Build Multi-Agent Automation Workflows

Learn how the Modal Context Protocol (MCP) connects large language models to databases, browsers, and local resources to automate end-to-end workflows with plain-English prompts.

Learn to build an agent that automates web browser tasks using Playwright or Selenium MCP servers, configured and started locally to run end-to-end automation with a large language model.

Build an agent that extracts data from a SQL database by framing complex queries using a MySQL MCP server, pip install, and the language model to orchestrate multi-agent workflows.

Build an agent that performs api testing and reads local files by wiring rest api and file system mcp servers, using postman collections to validate registration and login flows.

Demonstrates reading and writing to Excel sheets via an MCP server, building a single AI agent that can interact with Excel, browser, database, and APIs through MCP integration.

Compare single-agent architectures with multi-agent systems, showing how specialization enables parallel processing, easier maintenance, and scalable performance while reducing complexity and costs.

Explore sequential, parallel, hierarchical, and collaborative multi-agent workflows, coordinated by a manager agent, and understand agentic AI as autonomous systems powered by large language models and MCP tooling.

Download python from python.org and run Windows installer, enabling add python.exe to the path; ensure pip is installed. Verify with where python and python --version, then install PyCharm community edition.

Install Python on macOS, set up PyCharm Community Edition, and verify the Python 3 path with which Python 3 while noting the default Python 2 versus 3 locations.

Set up PyCharm and create a dedicated Python virtual environment for an Autogen project, install Autogen Agent Chat and MCP tooling, and configure OpenAI as the LLM.

Learn to define and run asynchronous Python functions using async and asyncio. Invoke these functions to execute code and start building foundational classes in Autogen for agentic ai workflows.

Explore three agentic workflows—single agent, multi-agent collaboration, and human-in-the-loop—using the Autogen framework. Build an assistant agent with OpenAI models and a basic brain setup.

Run an assistant agent in real time using run stream and view its streaming outputs in the console. Send text tasks or multimodal inputs (images or files) via multimodal message.

Learn to create two assistant agents with defined roles and coordinate their collaboration using a round robin group chat, enabling multi-agent problem solving.

Learn how termination conditions control multi-agent conversations in autogen, including max messages termination and message termination, with a practical walkthrough of a six-message stop mechanism.

Explore human-in-the-loop collaboration using the user proxy agent in AutoGen, enabling a human to chat with an AI agent in a round-robin group chat with termination via text cues.

Master the state saving mechanism for multi-agent workflows by saving and loading conversation context to JSON files, enabling seamless handoffs between agent one and agent two.

explains the selector group chat class that dynamically chooses which of three agents: researcher, writer, and critic acts next to research renewable energy trends and write an article.

Explore Autogen's multimodal web server, an inbuilt browser automation agent that launches Chromium and uses Playwright under the hood for web tasks.

Enable tooling support by wiring MCP workbench into the assistant agent, using MCP server parameters to perform file system actions and orchestrate multi-agent workflows within the Autogen framework.

Explore end-to-end multi-agent automation with an assistant and human in the loop to solve algebra problems, create teaching files, and store explanations using a file system.

Learn to build a two-agent workflow that autonomously fetches recent Jira bugs, consolidates a smoke test flow, converts it into Playwright automation, and reports results.

Sign up for Jira Cloud or Jira server, choose a scrum template, and create a credit card banking project to practice agile workflows in Jira.

Create clear Jira bugs with precise descriptions and reproduction steps, then design smoke tests for automated workflows using agentic ai, all in plain English.

Learn how to add Jira environment variables to an assistant agent using a dockerized MCP setup, including Jira URL, username, and API token, in a practical demo.

Build a Jira agent for multi-agent automation by configuring Jira URL, username, and token via environment variables, creating a Jira workbench, and scaffolding bug analyst and Playwright MCP server integration.

Learn how to build a browser automation agent by integrating Playwright with the MCP workbench, configuring async agents for Jira and automation tasks, and coordinating them in a round-robin multi-agent workflow.

Set up role-specific context for Jira bug analysis and Playwright automation, guiding multi-agent workflows with precise system messages, explicit steps, and synchronization to produce reliable smoke tests.

Develop a multi-agent workflow using round-robin teams, featuring a bug analyst and automation agent integrated with Jira and Playwright, driven by a generic agent factory and context-driven system messages.

Demonstrate agentic ai in action by coordinating bug analyst and automation agents to read jira bugs, design a smoke test, and validate end-to-end functionality with playwright.

Explore multi-agent automation workflows with separate database, API, and Excel agents using an agent factory to coordinate autonomously—register users, verify login, and write results to Excel.

Create a Python agent factory that builds three agents, configures MC configurations, and wires a database agent with a configurable model client and system message for coordinated workflows.

Build and connect an MCP config file to the test flow, initialize a MySQL workbench, and create a database agent for end-to-end connections. Prepare API and excel agents next.

Learn to build MCP configurations for rest API, file system, and Excel workbenches, create agents in the factory, and connect config files to the main test flow.

Build an end-to-end agentic ai workflow using an agent factory pattern to orchestrate database, api, and excel agents in a round-robin team with termination conditions and auto gen concepts.

Provide system messages that guide database, API, and Excel agents to retrieve registration data, construct validated API payloads, and coordinate end-to-end testing with unique emails and formatted fields.

Learn how to orchestrate Excel, API, and database agents with precise system messages and serial execution, signaling success and saving only registration data to Excel.

Run and debug a multi-agent workflow in the agentic AI framework, leveraging database, API, and Excel agents to achieve end-to-end automation with autonomous outputs.