A new comprehensive free course on GitHub is guiding learners through the full spectrum of AI engineering, starting from linear algebra and progressing all the way to building autonomous agents. The curriculum, titled "AI Engineering from Scratch," features 283 lessons organized into 20 phases, totaling approximately 320 hours of content.
Each lesson is designed to produce a real output, such as a prompt, skill, agent, or MCP server. The course is AI-native, meaning students learn by using Claude Code while studying. Instead of a certificate, the final deliverable is a functional toolkit.
The course has already attracted attention, with 147 views and 2 likes within three hours of being shared. It was announced by the YouTube channel @ziobuddalabs, which highlighted the course's hashtags including #AIEngineering, #LLM, #AgentAI, #ClaudeCode, #OpenSource, #MachineLearning, #AITools, and #DevAI.
Enthusiasts can access the repository at github.com/rohitg00/ai-engineering-from-scratch.