Our approach
AI literacy is not learning how to prompt a chatbot.
Applied AI education teaches learners how systems actually work: how a real problem becomes a governed workflow with AI inside it — and how to keep people accountable for what that system does.
Operating principles
Six principles behind everything we teach.
01
Problem before tools
Every engagement and every lesson starts with a named problem and the people who feel it. Tools are chosen last, not first.
02
Workflow before prompts
Prompting is one skill inside a larger discipline. Learners map the full workflow — inputs, decisions, approvals, actions — before writing a single instruction.
03
Humans stay responsible
Every system we teach has named human owners, explicit decision points, and approval gates for anything consequential. Responsibility is designed, not assumed.
04
Boundaries are a feature
Learners write the never-list before they build: what the system must never say, decide, send, spend, or store — and the safeguards that enforce it.
05
Evidence over enthusiasm
Usefulness is measured, not asserted. Every prototype carries a small set of honest metrics and a baseline from before the system existed.
06
Cost is part of design
Learners compute what a system spends — money, review time, attention — and compare it honestly against the value it creates.
The learning journey
From problem to improvement — one connected path.
This is a public educational abstraction of how applied AI systems are designed and operated. It teaches structure and responsibility without exposing any proprietary implementation.
01
Problem
Name a real problem and its cost
02
Context
Understand the environment
03
Workflow
Map the steps before tools
04
Data
Define allowed inputs and knowledge
05
Intelligence
Assign narrow AI jobs
06
Decision
Mark where choices matter
07
Human Approval
Design the review gate
08
Action
Act with limits and logs
09
Evidence
Record what happened
10
Learning
Study the evidence
11
Improvement
Change deliberately, with rollback
What learners practice
Twelve capabilities, exercised in every program.
These are the working skills of applied AI — each one practiced hands-on, not just described.
- 01Identify a real problem and the people affected
- 02Define the desired outcome in measurable terms
- 03Map the workflow before choosing tools
- 04Separate inputs, decisions, actions, and outputs
- 05Choose the right tools for each job
- 06Build a fast, cheap prototype
- 07Introduce human review and approval
- 08Measure cost, quality, risk, and usefulness
- 09Test failure conditions on purpose
- 10Improve the system from evidence
- 11Deploy responsibly and document what matters
- 12Explain what the system should never be allowed to do
What we don't do
Clear lines, kept.
No hype curriculum
No tool-of-the-week tours, no prompt trick collections, no promises of effortless automation. If a lesson doesn't help someone build or govern a real system, it doesn't ship.
No proprietary disclosure
All teaching uses public abstractions and sanitized examples. Private Crate South systems, client work, and internal architecture are never course material. See the IP boundary document in the repository.
No invented proof
We do not publish testimonials, partner logos, or outcomes that don't exist yet. Proof sections are honest placeholders until real work completes.
See the approach in motion.
The Applied AI System Map is the fastest way to understand how we teach — sixteen stages, five public use cases, every human responsibility marked.