AI Engineering From Scratch

Repository: github.com/rohitg00/ai-engineering-from-scratch
Stars: 30,486 ★ · Forks: 4,965
License: MIT
Languages: Python, TypeScript, Rust, Julia
Website: aiengineeringfromscratch.com
Author: Rohit G (also created Agent Memory)

Overview

“84% of students already use AI tools. Only 18% feel prepared to use them professionally. This curriculum closes that gap.”

A comprehensive, free, open-source AI engineering curriculum:

  • 503 lessons across 20 phases (~320 hours)
  • Every lesson ships a reusable artifact: prompt, skill, agent, MCP server
  • 150K+ readers, 241K page views in 30 days
  • Four languages: Python, TypeScript, Rust, Julia

Coverage

AreaTopics
AgentsAI agents, swarm intelligence, MCP
Deep LearningTransformers, attention, neural nets
LLMsFine-tuning, inference, alignment
Computer VisionCNNs, object detection, image gen
NLPTokenization, embeddings, RAG
Reinforcement LearningRLHF, policy gradients
EngineeringRust, TypeScript, production deployment

Topics (from repo)

agents · ai · ai-engineering · computer-vision · course · deep-learning · from-scratch · generative-ai · llm · machine-learning · mcp · nlp · python · reinforcement-learning · rust · swarm-intelligence · transformers · tutorial · typescript

Philosophy

“You don’t just learn AI. You build it. End-to-end. By hand.”

Most AI material teaches scattered pieces — a paper, a fine-tuning post, a flashy demo. This curriculum is the spine: 20 phases from foundations to production deployment.