Companies · 90 minutes
Executive briefing
For technical leaders who need to clarify AI use cases, risks and a realistic adoption plan for their team. Free.
Book the briefing● 21 September 2026 cohort · online · live · cohort under way
AI Developer in 13 weeks.
For developers. 24 live sessions. 23 seats.
In 13 weeks, your AI service is running in production: deployed, evaluated, measured. An applied AI engineering accelerator, not a certificate for the wall.
Free application. Response within 72 hours.
~/client-project $ npm run eval✓ 48/48 cases · faithfulness 0.93 · threshold 0.90 met~/client-project $ npm run break✗ faithfulness 0.71 · below threshold · deploy blocked~/client-project $
Start here
Six ways in.
Choose by profile and starting point: decide, frame and work with AI, or build and ship as a developer.
The full developer path, from Coding with AI through to the alumni programme, is laid out on one page. See the path
Work with AI
No code to write.
3 weeks · 6 sessions
AI at Work
For analysts, marketing, operations and administration: reliable AI workflows in your own job, verified, with measured time savings.
See the course4 weeks · 8 sessions
Leading an AI Project
For managers, product leads and founders: frame a use case, write the spec, read an evaluation, hold the budget.
See the course3 weeks or 1 day
AI and compliance
For legal, HR, compliance and the public sector: Law 25, PIPEDA, an internal usage policy and a risk register, ready to adopt.
See the courseBuild with AI
For people who already code.
Coding with AI · 5 weeks
AI-First Developer
For developers and teams who want to code with AI properly, inside their own codebase, without changing roles yet.
See Coding with AIThe Accelerator · 13 weeks
The Accelerator
For developers ready to build, evaluate and ship an AI feature in a real project, with measured proof.
See the AcceleratorWho it is for
Developers with at least one year of experience, in any language, who can reserve eight hours a week for thirteen weeks.
Not for absolute beginners, and not for aspiring machine-learning researchers: we train AI systems engineers, people who build, evaluate and deploy systems that use models, not people who train foundation models.
The system
Four pillars, one loop.
Everything ships.
Most training ends with a certificate. This one ends with a repository in production.
For the last four weeks you work in a team of three or four on a real organisation's problem, and each of you ships a feature you own, with your own evaluation score and your own contribution history.
What you show afterwards is not a certificate. It is a link.
« I teach what I ship to production. »
23 seats.
Starting 21 September 2026. Free application, response within 72 hours.