
Explore Autogen Studio, a low-code, open-source tool by Microsoft, designed for non-developers to rapidly prototype multi-agent applications via an interactive graphical user interface.
Discover how Autogen Studio provides a low-code interface to build and manage multi-agent solutions with agents, using skills, models, and workflows tested in the playground.
Install Autogen Studio via Anaconda by creating a Python 3.11 conda environment, configuring LM endpoints and OpenAI API keys, then launch the UI at localhost:8081.
Launch Autogen studio from a conda environment, access the ui on localhost port 8081, and explore the build and playground tabs for multi-agent workflows, sessions, and chat interactions.
Navigate the Autogen studio dashboard to build skills, models, agents, and workflows; create Python-based skills with secrets and database access, and configure and test OpenAI models.
Configure user proxy, assistant, and group chat agents in Autogen Studio by linking models and skills. Explore sequential and autonomous workflows and test interactions in the playground.
Explore how to create and save skills in Artisan Studio, build a primary AI assistant for detailed topic explanations, and generate PDFs from markdown via an autonomous workflow.
Autogen Studio enables a low-code workflow where a planning, language, and local AI agents collaborate in a travel group chat to generate itineraries with local events and language tips.
Configure Autogen Studio agents to connect with source models like llama 3.1 on your local machine via ollama endpoints. Test connections, link models, and compare performance to OpenAI's GPT-4.
Learn to install and launch Autogen Studio with Docker, building a Docker image from a Dockerfile and running a container on port 8081 to ensure safe, OS-level isolation.
Autogen Studio - Build Low Code Multi Agent AI Applications
AutoGen Studio is an open-source, low-code interface developed by teams at Microsoft to help in rapidly prototype AI agents, enhance them with skills, compose them into workflows and interact with them to accomplish tasks.
This course is a hand-on session to provide a deeper understanding on how multiple agents are linked into workflows to let a team of AI agents work for you to accomplish your tasks .
Course Overview :
1.Introduction to Autogen Studio
2.Getting Started With Autogen Studio
- Setting up environment (Conda + OpenAI)
- Navigating Studio Dashboard
3.Understanding Building Blocks of Autogen Studio
- Learn to create Skills and Configure Models
-Learn to create Agents , integrate them into Workflows and create sessions in Playground to execute Workflows
4.Advanced Techniques
- Walkthrough of Multi AI Agent workflow execution– Creating Skill
- Walkthrough of Multi AI Agent workflow execution – Generating PDF
- Walkthrough of Multi AI Agent workflow execution – Group chat with Multiple Agents to create Travel Itinerary
5.Using Open Source LLMs with Autogen Studio
-Step by Step instructions to link Agents to Llama3.1 model (Meta) running on local system, using Ollama tool
6.Launching Autogen Studio in Docker Container (Safe Code Execution)
-Step by Step instructions to install docker and launch Autogen Studio application in Docker Container