LEARNING
FDE Academy
Five modules on deploying AI inside somebody else’s business: reading a workflow before naming the problem, building the thing, putting it somewhere it is allowed to be, governing it, and handing it over. Every lesson ends with a task you carry out against a real deployment recipe, jurisdiction page or the Sprint structure on this site — not a quiz.
- Modules
- 5
- Lessons
- 24
- Cost
- Free
Free and self-paced. No certification is offered.
The learning path
5 modules · 24 lessons
In order, because each module assumes the one before it. Nothing is locked: start anywhere.
- 01Understand the customerWatch the work before naming the problem, cut it into parts one at a time, write down the constraints that decide the architecture, and scope one thing.workflow observationproblem decompositionconstraintsscoping4 lessons
- 02BuildWhat the model actually does, how you call it, why retrieval is the part that decides quality, where the state lives, and how the pieces are wired together.APIsLLM fundamentalsRAGworkflow automationdatabasesDocker5 lessons
- 03DeployRun a model on hardware you chose, decide where it lives, put a network and an identity in front of it, and leave behind logs somebody can use.cloudlocal LLMnetworkingauthenticationloggingmonitoring5 lessons
- 04GovernSpot the issues an AI deployment raises, tell a legal duty from a preference, read what a vendor has actually published, and leave a system somebody can supervise.privacysecuritydata residencyvendor assessmentAI governance5 lessons
- 05Forward deployFind who owns the process, work inside the team without becoming shadow IT, scope commercially in days and ranges, hand over properly, and turn the result into evidence.stakeholder managementworking inside customer teamshandoverdocumentationcommercial scoping5 lessons
01How to use this
Read a lesson, then do its task. The reading is the shorter half and it is not the point: the tasks are where the material becomes usable, and each one is carried out against something real on this site — a recipe with steps grounded in the project’s own documentation, a jurisdiction page with the instruments and their primary sources, or the day structure of the 3-Day AI Deployment Sprint.
Several tasks need hardware. Where a lesson needs a GPU it says so under What you need, and most of those can be done on a rented instance for an afternoon. Nothing here requires an account on this site, and nothing is submitted to us.
You can tick a lesson as done at the bottom of its page. The tick is stored in your browser, is not sent anywhere, and is not attached to anything — it is a bookmark for you, not a record for us.
02What this is not
This is written material with practical exercises, published free. It is not a course with an institution behind it, and stating that plainly is more useful to you than a page that leaves it ambiguous.
- There is no accreditation behind this material, from any body, in any jurisdiction.
- There is no certificate, badge, transcript or completion record. Nothing here is examined.
- No employer has agreed to recognise this material, and we have not asked one to.
- We do not place anyone in work, and finishing a module gives you no standing on this site.
- It is not registered under the Continuing Education Fund or any comparable scheme.
- It is not a professional standard, and it does not substitute for one where your work is regulated.
What can be earned on this site is a verified deployment — a record of a real system you built, checked by a person against evidence. That is not awarded for finishing a lesson, and it never will be.
03Where this is heading
Planned, not built. The loop below is the shape §11 describes; the two steps that would connect a lesson to a reviewed deployment do not exist yet, and no date is being offered for them. Nothing on this page depends on it.
- 01
LEARN
The five modules on this page, free and open, with no account.
IN THE PRODUCT TODAY
- 02
COMPLETE PRACTICAL PROJECT
Each lesson already ends in a task you can do against a real recipe. Nothing collects the result.
PLANNED
- 03
SUBMIT DEPLOYMENT
“I deployed this” on a stack or use-case page records what you built, privately by default.
IN THE PRODUCT TODAY
- 04
REVIEW
An administrator checks a submission against evidence before anything about it is published.
IN THE PRODUCT TODAY
- 05
VERIFIED SKILL / EXPERIENCE
Review currently verifies a deployment, not a skill. Nothing maps a finished lesson onto the skills taxonomy.
PLANNED
- 06
FDE PROFILE
A public profile carries verified deployments, skills, industries and jurisdictions.
IN THE PRODUCT TODAY
- 07
ELIGIBLE FOR CUSTOMER MATCHING
Enquiries are matched by a person against verified deployments — never against lessons read.
IN THE PRODUCT TODAY
04Where to go next
If you want to start with the engineering, module 02 leans on the twelve deployment recipes, and reading one end to end is a good half-hour before you begin. If you already deploy this kind of system for a living, module 05 and the deployment submission are the parts of this site that will be new.