I'm an Engineering Manager leading SRE and Platform teams. My day-to-day is reliability, infrastructure, and the systems that keep production running: which means I care about AI the way an operator does: what breaks, what it costs, and whether it holds up when the demo is over.
I built AI Academy because the content I needed didn't exist. Vendor docs are optimised to get you to their next API call. Academic papers assume you want a literature review. Blog posts are mostly surface-level hype or narrow tutorials that age in months. None of that helps you make real architectural decisions.
The goal here is practical and durable knowledge: the kind that lets you design a system, review a design, or call out a bad tradeoff. No vendor lock-in, no hand-waving, no prerequisites beyond curiosity and a willingness to read code.
Why AI Academy
Most AI education falls into two traps: hand-wavy overviews that leave you unable to make real decisions, or deep technical resources that assume you already understand the landscape. AI Academy sits in between.
Every module has three layers: a plain-English surface for anyone, a guided walkthrough with working code, and a deep-dive analysis for those who want the architecture and primary sources. You choose your depth.
The curriculum covers 9 tracks: from what an LLM actually is to running sovereign air-gapped deployments. Each module is tagged for volatility so you know what's stable knowledge and what needs re-reading in six months.
Open Source
AI Academy is fully open source. The curriculum content is licensed under CC BY-SA 4.0: you can use, adapt, and redistribute it as long as you attribute the source and share derivatives under the same licence. The site code is MIT.
If you spot an error, know a better explanation, or want to contribute a module, pull requests are open. See CONTRIBUTING.md for how to get started.
Stay updated
New modules, curriculum updates, and AI education content: occasionally, never spammy.
Newsletter coming soon.