Predicai Academy™
Build AI systems.
Not just prompts.
Five levels. Real artifacts. A human reviews your work before you advance.
Builder certification at Level 3. One payment. No subscription. Your capability doesn’t expire.
5
Levels
7
Quality gates
$149
One-time
Level 3
Certification
156 hrs
Annual capacity recovered
The Standard
Gold benchmarks, pass/fail criteria, and role examples for every level
The Standard is the enrolled student’s reference. It defines what good looks like, what gets rejected, and how Cai supports your build at each level.
Gold examples
Pass / revise rules
Role variants
Cai prompts
Submission flow
The five levels — click to expand
01
Translator
“You’re getting output. It’s just never quite right.”
Deliverable
Intent Protocol
prompt architecture
▼
“You open ChatGPT. You type something. It gives you something back. Sometimes it’s useful. Most of the time you rewrite it anyway. You don’t know why it works when it does.”
The Intent Protocol — You learn to use the R.I.S.E. Framework to architect every strategic request: Role, Input & Context, Steps, Expectation. The result is a reusable prompt structure that produces consistent output. A human Predicai architect reviews it against the 7 Gates and unlocks Level 2.
View Level 1 gold standard & pass criteria →
02
Operator
“You’re busy all week but nothing actually moves forward.”
Deliverable
Time Ledger
task audit
▼
“Every week you clear your inbox, attend your meetings, update your trackers. You go home tired. On Friday you realize nothing actually shipped. The same tasks came back on Monday.”
The Time Ledger — A 3-day audit to find your Zombie Tasks: high-repetition, low-judgment work that AI can own. You categorize every recurring task as Automate, Augment, or Human-Only. You calculate the annual cost of drudgery using your real hourly rate.
View Level 2 gold standard & pass criteria →
03
Builder
“You keep asking AI for complex work and it keeps failing.”
Deliverable
Agentic Blueprint
workflow design
✓ Certification
“You paste a 50-page document and say ‘write a strategy’. It gives you something that sounds good but isn’t. You ask for complex analysis. It collapses halfway through. You don’t know why.”
The Agentic Blueprint — You stop prompting and start building assembly lines. Every cognitive workflow is three steps: Reader (ingest), Thinker (reason), Writer (output). You map your most expensive Zombie Task into this chain, run it in isolation, verify each output. Passing Level 3 earns the Predicai Builder Certification.
View Level 3 gold standard & certification criteria →
04
Forecaster
“You didn’t see the risk coming until it was too late.”
Deliverable
Signal Map
risk radar
▼
“Most leaders read autopsies — reports on what happened. Level 4 teaches you to build radar: reports on what will happen. The shift from reacting to data to seeing the curve.”
The Signal Map — Use AI to scan data for Weak Signals — trends invisible to the naked eye. Build scenario models across Linear, Exponential, and Black Swan futures. Submit a Human-in-the-Loop Protocol answering: at what confidence threshold does the AI stop? Who monitors it? What is the rollback?
View Level 4 gold standard & pass criteria →
05
Leader
“Your team is already using AI. Just badly.”
Deliverable
Strategic Doctrine
team AI policy
▼
“If you don’t design the system, your team builds their own — badly. Shadow AI is already running in your organization. The question is whether it has guardrails.”
The Strategic Doctrine — Define the rules of engagement for your team’s AI use: Intake, Governance, Enablement. Identify the Red Lines — what AI must never do autonomously. Final requirement: teach the workflow to one colleague. If you can teach it, you have scaled it.
View Level 5 gold standard & pass criteria →
Enrolled students
Need the Gold Standard benchmarks, role examples, or submission rubrics?
The System Dossier
You don’t graduate with a certificate. You graduate with proof.
All five artifacts assembled into one cohesive portfolio. A Business Case slide with real ROI math. A System Logic map showing how all five levels connect.
A pile of car parts is not a car. The Dossier is the engine. Show the connections. Prove the system holds up.
System Dossier — five components
L1Intent ProtocolPDF
L2Time LedgerDB + ROI
L3Agentic BlueprintCertified ✓
L4Signal Map + HITLDashboard
L5Strategic DoctrinePolicy
Required: Business Case
“[Hrs/wk] × [Rate] = Annual capacity recovered.”
What to expect
Common questions, answered directly.
Is this for beginners or advanced users?
Both. Level 1 starts from first principles — no prior AI experience needed. Level 5 is designed for people already running AI workflows who want governance and team systems. Most people enter at Level 1 or 2 and find the structure immediately useful regardless of where they started.
How long does it take?
It is self-paced. Each level takes roughly 2–4 hours of focused work. Most people complete all five levels in 3–6 weeks. The pace is yours. The review step happens when you submit — not on a schedule.
What tools do I need?
Any major AI model works — ChatGPT, Claude, Gemini. You do not need to buy new tools. Academy teaches the framework that makes any AI tool more capable in your hands, not dependency on a specific platform.
What does human review actually mean?
A Predicai architect reviews your submitted artifact against the 7 Gates of Quality. This is not grading — it is verification. They confirm the system is real, the logic holds, and the output would survive scrutiny in a real work context. Then they unlock your next level. Full review criteria are published in The Standard.
Is this relevant for work, consulting, or leadership?
All three. Levels 1–2 are primarily individual capability. Level 3 is where workflow design begins. Levels 4–5 are explicitly for people making decisions that affect others — leaders, consultants, and operators who need a framework their team can adopt, not just personal techniques.
Where do I find the pass/fail criteria and Gold Standard examples?
All benchmarks, role variants, task sheets, and Cai prompts live in The Standard. It is the enrolled student’s operating reference. Every level has a Gold, Good, and Bad example so you know exactly what to aim for before you submit.
Who Academy is for
If any of these sound like you, you are in the right place.
Knowledge workers
You use AI every week but the output is inconsistent. You know it could do more. You just don’t have a system for making that happen reliably.
Consultants & operators
You are billing time on work AI could automate. You need a structured way to build AI into client deliverables without sacrificing quality or accountability.
Leaders & managers
Your team is already using AI. Nobody agreed on how. You need a framework to govern it, teach it, and make it produce consistent results across the organization.
Self-directed learners
You are self-taught and progressing. What you need is not more content. You need a structure that forces you to produce real outputs and get them reviewed.
What human review means
A Predicai architect reviews your artifact against the 7 Gates of Quality before unlocking the next level. Not grading. Verifying the system is real and holds up.
What you build
Five artifacts applied to one real workflow from your job. Not exercises. Not simulations. Real work, reviewed by a human, assembled into a System Dossier you can show.
Not sure which level to start at?
Take the Capability Scorecard. It places you on the capability ladder in under 5 minutes and tells you exactly where to enter. No account required.
$149 one-time · No subscription · Lifetime access
Build your
five-artifact dossier.
Start at Level 1. Work through the system. Graduate with proof, not just a certificate.