Coming Soon

Build AI
Agents .

Build 4 production AI agents from scratch — a personal assistant with RAG, an MCP tool server, multi-agent trip planners, and an AI career coach. Hands-on Arabic course. No fluff, just real code.

Pre-Enroll Now → See the Projects ↓
agent.js
import { createAgent } from "langchain"; const agent = createAgent({   model,   tools: [searchKnowledgeBase],   memory,   systemPrompt: `You are a helpful     assistant. Search the knowledge     base. Be accurate.`, })
LangChainLangChain LangGraphLangGraph LangSmithLangSmith OpenAIOpenAI OllamaOllama MCPMCP A2AA2A ReactReact Node.jsNode.js PineconePinecone AnthropicAnthropic
What You'll Build

4 Real Projects.
Production Code.

Each project is a complete full-stack application with a React frontend, Node.js backend, and real AI integrations. You'll build, run, and own every line of code.

Personal Assistant → MCP Server → Trip Planner → Career Assistant
01
AI Personal Assistant
Single agent · 3 modes · RAG pipeline

Build a full-stack AI assistant that ingests PDF documents into a Pinecone vector database and answers questions using RAG. Switch between three modes — RAG for your own documents, API for live web search, and MCP for tool-based search — all through a polished chat interface.

  • PDF ingestion and text chunking
  • Vector embeddings with Pinecone
  • ReAct agents with LangChain
  • OpenAI and Ollama integration
LangChain OpenAI Ollama Pinecone MCP React
02
MCP Search Server
Tool server · Streamable HTTP · stdio

Build your own MCP-compliant tool server that exposes web search and image search as tools any LLM client can use. Implement both Streamable HTTP and stdio transports, and understand when to use MCP vs REST APIs vs direct SDK calls.

  • MCP protocol deep dive
  • Streamable HTTP vs stdio transports
  • Building reusable tool servers
  • MCP vs REST vs SDK trade-offs
MCP SDK Node.js Express
03
AI Trip Planner
Multi-agent · LangGraph + A2A · built twice

Build the same multi-agent trip planning system two different ways. First with LangGraph — orchestrating search, budget, and itinerary agents through a centralized state graph with real-time SSE streaming. Then rebuild it with the A2A protocol — where each agent runs as an independent service communicating via JSON-RPC. Same problem, two architectures, so you know exactly when to use each.

  • LangGraph state machines
  • A2A protocol and agent discovery
  • Parallel agent orchestration
  • SSE streaming and JSON-RPC
LangGraph A2A MCP Claude React
04
AI Career Assistant
Multi-agent · LangGraph + Anthropic Claude

Build a multi-agent career analysis system powered by Anthropic's Claude. Three specialized agents — resume analyzer, market researcher, and gap analyst — work together through a LangGraph pipeline to deliver personalized career recommendations with real-time progress streaming.

  • Anthropic Claude integration
  • Specialized agent pipelines
  • Resume parsing and analysis
  • Real-time SSE progress updates
Anthropic LangGraph Node.js React
Topics Covered

Everything You Need to Know.

RAG
Retrieval-Augmented Generation with vector search
MCP
Model Context Protocol, transports, and tool servers
Multi-Agent
Orchestration, state graphs, and parallel agents
A2A Protocol
Distributed agents with discovery and JSON-RPC
Local LLMs
Run models offline with Ollama
Full-Stack
React + Node.js + real deployable apps
Early Access

Reserve Your Spot.

Drop your email and you'll be the first to know when the course launches — including early-bird pricing.

4 projects · Full source code · The complete AI agents toolkit
✓
You're on the list!
We'll notify you the moment the course goes live.