Skip to content

Learn to build with AI, properly.

Practical AI and automation training. For company teams, run on your own processes and tools. For students and colleges, run as project-based cohorts where you ship something real.

Coming soon

While you wait

Ten courses we would put in front of you anyway.

Academy is not open yet, so here is what we would tell you to watch in the meantime. Every one is free, every one runs for at least three and a half hours, and none of them is ours. Play them here or open them at the source — nothing loads from YouTube until you press play.

10 courses · 171 hours · nothing to sign up for

  • The best starting point on this list. A real university course, taught properly, from search algorithms through to neural networks.

    • Search
    • Knowledge
    • Optimisation
    • Neural networks
    • Python
  • The longest thing here by a distance. Treat it as a reference you work through over months, not a weekend.

    • Computer vision
    • TensorFlow
    • CNNs
    • Transformers
    • Deployment
    37h 17mFree to watch · no published licenceSource ↗
  • Generative AI Full Course – Gemini Pro, OpenAI, Llama, Langchain, Pinecone, Vector Databases & More

    freeCodeCamp.org · Krish Naik, Sunny Savita and Boktiar Ahmed Bappy

    Broad coverage of the current generative stack. Useful if you need to know what all the pieces are before choosing any of them.

    • LLMs
    • LangChain
    • Vector databases
    • RAG
    • Fine-tuning
    30h 18mFree to watch · no published licenceSource ↗
  • The one to watch if you intend to build rather than to understand. Code-first the whole way through.

    • PyTorch
    • Tensors
    • Neural networks
    • Transfer learning
    • Python
  • Agents, end to end, including the parts everyone skips — memory, human-in-the-loop and actually deploying the thing.

    • LangGraph
    • LangChain
    • Multi-agent
    • RAG
    • Deployment
    24h 16mFree to watch · no published licenceSource ↗
  • GenAI Essentials – Full Course for Beginners

    freeCodeCamp.org · Andrew Brown

    Aimed at people who have to make decisions about generative AI rather than write it. Strong on the cloud side.

    • Generative AI
    • Prompting
    • Cloud
    • Model deployment
    • RAG
    22h 54mFree to watch · no published licenceSource ↗
  • A calm, well-paced introduction. If the longer courses feel like too much to commit to, start here.

    • TensorFlow
    • Neural networks
    • Computer vision
    • NLP
    • Python
    6h 52mFree to watch · no published licenceSource ↗
  • Let's reproduce GPT-2 (124M)

    Andrej Karpathy · Andrej Karpathy

    Four hours building GPT-2 from an empty file. Nothing else on this list will teach you as much about what a language model actually is.

    • Transformers
    • GPT-2
    • PyTorch
    • Training
    • From scratch
    4h 01mFree to watch · no published licenceSource ↗
  • Genuinely for beginners, and the shortest real commitment here. Good for a team that needs shared vocabulary before a project starts.

    • Supervised learning
    • Classification
    • Regression
    • TensorFlow
    3h 54mFree to watch · no published licenceSource ↗
  • Deep Dive into LLMs like ChatGPT

    Andrej Karpathy · Andrej Karpathy

    The clearest explanation available of how these models are built and why they behave as they do. Watch this before you commission anything with an LLM in it.

    • LLMs
    • Pre-training
    • Fine-tuning
    • RLHF
    • Hallucination
    3h 31mFree to watch · no published licenceSource ↗

None of this is our work. Every course belongs to the person or organisation credited on its card, and each title links to the source so you can confirm that for yourself. We have not edited, re-uploaded or mirrored any of it. Where a course carries a published licence we name it; where it does not, it is free to watch but not openly licensed, and we say nothing more than that.