Udemy
    •  
    •  
    •  
    •  
    •  
    •  
    •  
    •  
Subscribe
Turn what you know into an opportunity and reach millions around the world.
Learn More
Your cart is empty.
Keep shopping
Build AI Agents: CrewAI, LangGraph, OpenAI SDK, Autogen, MCP
New
Rating: 4.7 out of 5(2 ratings)
5 students

Build AI Agents: CrewAI, LangGraph, OpenAI SDK, Autogen, MCP

End-to-end AI agent bootcamp - from your first LLM call to multi-agent systems you can deploy at work
New
Created byDeesa Technologies
Last updated 7/2026
English

What you'll learn

  • LLM fundamentals: tokens, cost, latency, capabilities & limitations
  • Prompt engineering that controls output (JSON, classification, extraction)
  • Agent design patterns: tool use, multi-step reasoning, failure modes
  • OpenAI Agents SDK for research copilots, meeting notes → action items, support bots
  • CrewAI multi-agent orchestration for software teams, product launches & hiring
  • LangGraph stateful workflows with human-in-the-loop approval
  • AutoGen collaborative code review and problem-solving agents
  • MCP integration for databases, PDFs, and live market data
  • Guardrails to prevent unsafe or unreliable AI outputs
  • Observability: logging, tracing, failure simulation & quality evaluation
  • Deployment: FastAPI, webhooks, cost optimization

Course content

13 sections • 78 lectures • 23h 50m total length

  • What This Course Is About — Goals, Structure & What You'll Build11:31
  • Access the Full Course Code on GitHub — Clone, Navigate & Use the Repo0:02
  • Install Required Software — Python, UV, IDE & Project Dependencies21:55
  • Set Up OpenAI & DeepSeek API Keys — Create, Store & Test Your Keys24:39

Requirements

  • Basic Python knowledge
  • Curiosity and willingness to build — no ML PhD required

Description

Stop watching AI demos. Start building AI systems companies actually ship.


This hands-on bootcamp takes you from your first LLM API call to production-grade multi-agent systems — using the tools shaping AI engineering today: OpenAI Agents SDK, CrewAI, LangGraph, AutoGen, and MCP.

You won’t just learn theory. You’ll design, build, debug, secure, and deploy real AI applications in Python.


What you’ll build:

  • 29 exciting AI projects that we will design from scratch

  • AI research assistant, email generator & chatbot (Module 0)

  • Prompt playground, JSON formatter & email classifier

  • Resume analyzer, data extractor & SQL assistant

  • Web search agent & multi-step task agent

  • OpenAI SDK: research copilot, meeting notes → action items, support bot

  • CrewAI flagship: AI software engineering team (Developer → Reviewer → Tester)

  • LangGraph customer support & human-in-the-loop approval workflows

  • AutoGen collaborative code review & problem-solving agents

  • MCP: internal data assistant, knowledge base agent, trading simulation

  • Debug dashboard, failure simulator & output evaluator

  • GitHub AI PR code reviewer (FastAPI + webhooks)


This course doesn't just teach you how to design and build AI systems — it also shows you how and when AI fails, and what to do about it. It's a production-ready bootcamp that takes you all the way to deployment, including a GitHub AI PR reviewer you can run in real time and watch interact with live pull requests.


By Deesa Technologies — production-focused courses built by engineers who ship real systems.

Enroll now and build your first AI agent in Module 0.

Who this course is for:

  • Python developers who want to move into AI engineering
  • Software engineers building internal AI tools at work
  • Data professionals who need practical agent workflows, not theory-only content
  • Career switchers targeting AI Engineer / ML Engineer / Agent Developer roles
  • Anyone tired of copy-paste ChatGPT tutorials who wants real system design skills

Report abuse