Do I need to have used AutoGen before?

No. You will set it up from scratch. The course assumes you have built at least one tool-calling agent with any framework (OpenAI SDK, LangChain, or a raw loop).

Which LLM provider do I need?

Any OpenAI-compatible endpoint works: OpenAI, OpenRouter, Fireworks, or local models via LiteLLM. Gemini is supported through a dedicated client.

How is this different from LangGraph?

AutoGen is conversation-first: agents talk to each other in a GroupChat and the framework handles the turn-taking. LangGraph is graph-first: you define explicit state transitions. Same goal, different mental model.

Is the travel support scenario realistic?

Yes. The same pattern powers airline support bots, insurance claim flows, and internal IT helpdesks. Specialist agents plus a supervisor is how production systems scale.

Premium course

Stop wrestling one mega-agent. Orchestrate a team.

Name: Multi-agent systems with AutoGen GroupChat
Price: 49 USD
Availability: InStock

Cramming every tool into one system prompt collapses the moment the task gets interesting. In this workshop you build a travel support assistant with AutoGen GroupChat, split the work across specialist agents, and route traffic through a supervisor that actually knows who should answer.

Enroll Preview curriculum

Still deciding? Ask first.

Message a mentor about fit, prerequisites, or where to start. Replies come on WhatsApp, usually within a day.

Curriculum fit, prerequisites, or where to start
Honest answer, no pressure to enroll

Engineers are learning here from

NVIDIAMICROSOFTGRABWISEPIPEDRIVEBOLTGLIA

Build a production-grade travel support assistant with AutoGen. Wire up tool-calling agents, orchestrate them with GroupChat, and route work through a supervisor so specialist agents collaborate instead of collide.

Orchestrate specialist agents with AutoGen GroupChat and a supervisor router.

What you'll ship

Real projects, not toy demos.

A travel support assistant with multiple specialist agents
Read tools for lookup, search, and flight status
Write tools for booking, cancellation, and taxi reservations
A supervisor that routes each user message to the right agent
Streaming FastAPI endpoints that surface tool calls and reasoning

What you'll learn

You finish able to:

Set up AutoGen with a streaming model client and tool-calling agents
Wire read tools that let an agent see booking, hotel, and flight data
Wire write tools that let an agent book, cancel, and modify state safely
Wrap agents behind a FastAPI streaming endpoint that emits tool events
Design a supervisor that routes user intent to the right specialist agent
Extend the system with new tools without destabilising existing behaviour

Curriculum

From a single AutoGen agent to a supervisor-routed team.

01
AutoGen basics
Install AutoGen, wire up a streaming model client, and run your first agent with a tool
3 lessons
02
Tools and state
Expand the toolbox with read and write tools, then wrap the agent behind a streaming FastAPI endpoint
3 lessons
03
Multi-agent orchestration
Split tools across specialist agents, add a supervisor, and extend the system with new capabilities
3 lessons

Who it's for

Is this for you?

Engineers with a single-agent prototype

who are tripping over tool-calling chaos and want real orchestration

Backend developers shipping AI features

who need structured coordination instead of one bloated system prompt

AI engineers comparing frameworks

who want to see how AutoGen GroupChat actually works end to end

FAQ

Common questions.

Do I need to have used AutoGen before?
No. You will set it up from scratch. The course assumes you have built at least one tool-calling agent with any framework (OpenAI SDK, LangChain, or a raw loop).
Which LLM provider do I need?
Any OpenAI-compatible endpoint works: OpenAI, OpenRouter, Fireworks, or local models via LiteLLM. Gemini is supported through a dedicated client.
How is this different from LangGraph?
AutoGen is conversation-first: agents talk to each other in a GroupChat and the framework handles the turn-taking. LangGraph is graph-first: you define explicit state transitions. Same goal, different mental model.
Is the travel support scenario realistic?
Yes. The same pattern powers airline support bots, insurance claim flows, and internal IT helpdesks. Specialist agents plus a supervisor is how production systems scale.

Pricing

Pick the path that fits.

Subscribe to Pro for every paid course, or buy just this one.

Subscribe to Pro

Recommended

Unlock this course and every paid course plus workshop replays. One subscription.

$35$16/mo · Billed annually

You save 54% with regional pricing

Every paid course unlocked
All workshop replays included
New releases as they ship
Cancel anytime

Buy this course

One-time purchase. Lifetime access to every lesson, exercise, and update.

$49$29one-time

You save 41% with regional pricing

Lifetime access to every lesson and future update
Hands-on projects you can show in interviews
Certificate when you finish

Still deciding? Ask Param a question

After this workshop:

Multi-agent is not a buzzword. It is how serious AI systems are built. Start here.

Enroll

Multi-agent systems with AutoGen GroupChat

$29 one-time

Enroll

Premium course

Stop wrestling one mega-agent. Orchestrate a team.

Enroll Preview curriculum

Still deciding? Ask first.

Message a mentor about fit, prerequisites, or where to start. Replies come on WhatsApp, usually within a day.

Curriculum fit, prerequisites, or where to start
Honest answer, no pressure to enroll

Engineers are learning here from

NVIDIAMICROSOFTGRABWISEPIPEDRIVEBOLTGLIA

Orchestrate specialist agents with AutoGen GroupChat and a supervisor router.

What you'll ship

Real projects, not toy demos.

A travel support assistant with multiple specialist agents
Read tools for lookup, search, and flight status
Write tools for booking, cancellation, and taxi reservations
A supervisor that routes each user message to the right agent
Streaming FastAPI endpoints that surface tool calls and reasoning

What you'll learn

You finish able to:

Set up AutoGen with a streaming model client and tool-calling agents
Wire read tools that let an agent see booking, hotel, and flight data
Wire write tools that let an agent book, cancel, and modify state safely
Wrap agents behind a FastAPI streaming endpoint that emits tool events
Design a supervisor that routes user intent to the right specialist agent
Extend the system with new tools without destabilising existing behaviour

Curriculum

From a single AutoGen agent to a supervisor-routed team.

01
AutoGen basics
Install AutoGen, wire up a streaming model client, and run your first agent with a tool
3 lessons
02
Tools and state
Expand the toolbox with read and write tools, then wrap the agent behind a streaming FastAPI endpoint
3 lessons
03
Multi-agent orchestration
Split tools across specialist agents, add a supervisor, and extend the system with new capabilities
3 lessons

Who it's for

Is this for you?

Engineers with a single-agent prototype

who are tripping over tool-calling chaos and want real orchestration

Backend developers shipping AI features

who need structured coordination instead of one bloated system prompt

AI engineers comparing frameworks

who want to see how AutoGen GroupChat actually works end to end

FAQ

Common questions.

Do I need to have used AutoGen before?
No. You will set it up from scratch. The course assumes you have built at least one tool-calling agent with any framework (OpenAI SDK, LangChain, or a raw loop).
Which LLM provider do I need?
Any OpenAI-compatible endpoint works: OpenAI, OpenRouter, Fireworks, or local models via LiteLLM. Gemini is supported through a dedicated client.
How is this different from LangGraph?
AutoGen is conversation-first: agents talk to each other in a GroupChat and the framework handles the turn-taking. LangGraph is graph-first: you define explicit state transitions. Same goal, different mental model.
Is the travel support scenario realistic?
Yes. The same pattern powers airline support bots, insurance claim flows, and internal IT helpdesks. Specialist agents plus a supervisor is how production systems scale.

Pricing

Pick the path that fits.

Subscribe to Pro for every paid course, or buy just this one.

Subscribe to Pro

Recommended

Unlock this course and every paid course plus workshop replays. One subscription.

$35$16/mo · Billed annually

You save 54% with regional pricing

Every paid course unlocked
All workshop replays included
New releases as they ship
Cancel anytime

Buy this course

One-time purchase. Lifetime access to every lesson, exercise, and update.

$49$29one-time

You save 41% with regional pricing

Lifetime access to every lesson and future update
Hands-on projects you can show in interviews
Certificate when you finish

Still deciding? Ask Param a question

After this workshop:

Multi-agent is not a buzzword. It is how serious AI systems are built. Start here.

Enroll

Multi-agent systems with AutoGen GroupChat

$29 one-time

Enroll