Your cart is currently empty!
CAID Console
CAID · Certificate in AI Direction — Program Console
The AI Direction Program · “Direct AI. Multiply your impact.”
Using AI → Directing AI → Architecting AI — three peer competency areas, not a ladder. This program owns the middle: the universal one. Six Domains · Eight Direction Abilities · a growing library of reusable Skills.
What will I experience?
~12 minthe diagnostic
minutesreport arrives
same daystart Foundation
< 4 weeksmax wait for a module
~3 h/weekmodule load
~1 quarterto CAID (fastest ~6 weeks)
1
An invitation — at your certification workshop, or an email about your work, not about AI.
2
Twelve questions about your work — about 12 minutes. A mirror, not a test: no score, no pass or fail. Only one question even mentions AI.
3
Your AI Direction Report arrives within minutes — a personal map of your week: where your hours go, how many are worth reclaiming, which 2–3 modules fit you (explained in your own words), and reusable AI Skills suggested for your job.
4
A live session (optional) — monthly, for report holders. Watch AI directed live on real work like yours; ask anything.
5
Enroll any day, start the same day — Foundation (AI Direction Essentials) is self-paced, about two weeks at your rhythm: the Eight Direction Abilities — Delegate, Frame, Context, Plan, Steer, Validate, Land, Capture — taught as an interacting loop, with AI feedback on every exercise.
6
Your module begins — new cohorts start the first weekend of every month. Maximum wait under 4 weeks, typically 1–2, often none after Foundation.
7
Four weekends, about 3 hours a week — one 2-hour weekend session (live demonstration, then critique of real participant work); weekdays you apply it to your own job. Your homework is your work. AI feedback on every submission, within minutes.
8
Week 4: you demonstrate — a real deliverable from your own job, produced by directing AI: a walkthrough plus a 10-minute conversation. Every viva ends the same way: “When should this result not be used?” Not there yet? Precise feedback and one retry.
9
Module badge earned — your named Skills and deliverables join your capability profile. Choose the next module (or run two in parallel if your schedule allows).
10
Certificate in AI Direction — earned with Foundation plus two demonstrated domains: at least one capstone in real workplace use, and at least one of your Skills successfully run cold by someone else. A live, shareable capability profile: the domains you demonstrated, the Skills you built, the portfolio you can show anyone. Further modules add endorsements. Drawn to building AI solutions? CAISA is your next doorway — a different craft, not a higher rank.
The twelve questions, as you will see them
- What is your role, and what does your team or organization most depend on you to deliver, improve, or decide?
- Think of your last full work week — and any less frequent but important work from a typical month. What three or four activities consumed most of your time, roughly how many hours did each take, which carried the greatest consequences if done poorly or late — and which would you most like to permanently shrink or delete?
- How much of your work involves getting on top of unfamiliar or fast-moving information — researching, reading long documents, comparing evidence, keeping current? Describe the most recent example, what the answer was needed for, and what made it difficult.
- What data, if any, passes through your hands? What must you do with it — define or prepare it, find patterns, compare performance, forecast, visualize, or explain implications? Describe one recent analysis and the decision or action it supported.
- What problems or decisions do you own or materially influence in a typical month? Describe one recent example — and what made it hard: unclear framing, uncertain causes, limited evidence, competing options, risk, or stakeholder disagreement?
- What do you need others to understand, trust, decide, or do? Think of reports, proposals, presentations, training, meetings, or difficult messages. Which communication takes longest or matters most — and which audience is hardest to reach?
- Which recurring piece of work moves through similar steps, handoffs, reviews, or approvals each time? How often does it occur, how long does it take — and where do delays, rework, dropped actions, or exceptions arise?
- Is any work currently held together by spreadsheets, forms, email, or WhatsApp that more than one person must enter, track, approve, view, or update? What should a useful small tool do, who would use it — and would you be willing to help build, test, and own it?
- Which parts of your work are difficult to hand over because the necessary context, standards, or judgment live mostly in your head? What could become reusable if that context were captured?
- Looking at the next six to eight weeks, which piece of upcoming work matters most to get right — a question to answer, data to analyze, a decision to make, something important to communicate, a recurring process to improve, or something you need to track? Who would use or review the result?
- How do you currently use AI tools? What has worked well — and where have they disappointed, misled, or worried you?
- Looking ahead two or three years, what would you like to be known for professionally? Choose the two improvements that would help most right now — getting on top of information · drawing insight from data · solving problems and making decisions · communicating and influencing · redesigning recurring work · building small tools — and say why those two.
Your answers are used only to generate your personal report. Describe your work in general terms — no confidential figures, client names, or personal data are needed. If an area is barely part of your work, say so; that is useful information.
How do we fill it?
The funnel, with conversion checkpoints
| Stage | What happens | Checkpoint metric |
|---|---|---|
| Wednesday AI workshop | Fresh GB/BB graduates. Restructured: open with the diagnostic themes, demonstrate AI direction live on participants’ real answers, close with the assessment link. Bundled with certification as the closing event — never a gate. | attendance → diagnostic completion |
| Alumni mailing | Whole database: GB, BB, MBB, CAISA. Segmented sends, job-first copy, tested on slices before rollout. | open → diagnostic completion |
| Diagnostic | 12 questions, self-serve, automated. Qualification is behavioral: completing it is the filter. Nothing to game. | start → finish rate |
| AI Direction Report | Personalized within minutes; quotes the respondent verbatim; every number traceable to their answers. The report is the first demonstration of directed AI. | report → live-session registration |
| Live session | Monthly, report holders only. Depth and conversion, never awareness. | session → enrollment |
| Enrollment | Always open; Foundation starts same day. | module queue lengths (drives the calendar) |
Segments and message angles
| Segment | Channel | Angle |
|---|---|---|
| Fresh GB/BB (passing out) | Wednesday workshop | Your certificate is the occasion; your job is the subject. Their ChatGPT habit is qualification, not disqualification — convert extractive use into skilled direction. |
| Recent alumni | “Get hours of your week back.” Job pain first; program name second. | |
| Senior alumni / MBB | Leadership leverage; team-level direction. The diagnostic may sort them toward CAISA. | |
| CAISA alumni | Directing for your own productivity — a peer skill, not a step down. Non-hierarchy made operational. |
Positioning spine
- Category creation: nobody searches “AI direction course” — the first market is a database we own, so we can afford to define the category, as with CAISA.
- The film-director metaphor: a director does not act or operate the camera, yet the film is theirs. You don’t need to build AI — you need to direct it.
- Never tools training: organized around how professionals create value (learn · analyze · decide · communicate · systematize · build). Tools churn quarterly; these activities do not.
- The credential is an identity: the Certificate in AI Direction answers “what do you want to be known for?” — backed by demonstrated real work, a LinkedIn headline rather than an attendance certificate.
- Trust rule: the report’s numbers are conservative, ranged, and traceable to the respondent’s own words. Never invented precision.
- The three layers, one line each: Domains show where you can direct AI. Abilities show how you direct it. Skills make successful direction repeatable.
How does it run?
The weekend engine — 8 facilitator hours
| Slot | Assignment | Notes |
|---|---|---|
| Sunday h1–2 | Track A permanent high-demand module | New cohort monthly. Likely Communication & Influence — let diagnostic data decide. |
| Sunday h3–4 | Track B second high-demand module | New cohort monthly. Likely Workflow & Productivity Systems. |
| Saturday h1–2 | Track C rotating slot | Remaining modules take turns by queue length. Quarterly floor date each; extra run when queue reaches 10. |
| Saturday h3–4 | Program Slot | Wk1 live funnel session · Wk2 capstone overflow + resubmission vivas · Wk3 open clinic (alumni welcome; Skill cold-run transfer tests happen here) · Wk4 buffer/rest. |
Inside every module track, the same month: Wk1 kickoff and demonstration · Wk2–3 critique of direction trails · Wk4 capstone block with vivas. Weekday layer runs at zero facilitator hours (diagnostic, report, Foundation, AI evaluator). Only weekday commitment: ~2 hours Friday prep.
Capacity arithmetic: one cohort = one slot × 4 weeks = 8 live hours. Steady state: 3 cohorts/month at 10–20 participants → 30–60 module enrollments monthly from 8 weekend + 2 prep hours. Cohort cap set by vivas: ~10 per capstone block, ~20 with overflow. Participant clock: Foundation (~2 wks, self-paced) + wait (<4 wks) + two modules → CAID in about a quarter; parallel A+B possible → ~6 weeks.
Calendar rules
- Enrollment always open; Foundation absorbs all waiting.
- Module cohorts start the first weekend of every month.
- Track C: published quarterly floor per module; queue of 10 triggers an extra run.
- Launch mode (months 1–3): Tracks A and B only; Track C dark while queues fill; Program Slot carries funnel + overflow. ~4–6 h/week while content is recorded and the AI evaluator is tuned on real submissions. During pilots: keep 2–3 scored capstones per module as anchor exemplars — the calibration standard being written down while the founder is still the only assessor.
- Scale path: demand beyond four slots → CAID coaches inherit tracks one slot at a time. Founder permanently keeps Problem Solving & Decision Making. Pre-delegation rule: anchor exemplars and the Friday moderation routine must exist before the first track is handed over.
Systems inventory
| System | Status |
|---|---|
| Diagnostic form (12 questions) | To build from the diagnostic v2 spec; pilot-time on alumni before launch (the ~12-min promise and the instrument must never disagree) |
| AI advisor — AI Direction Report generator | To build — routing rules, collision rules, and guardrails are specified in the diagnostic v2 spec; output drafting remains, sibling to module anchors |
| Module behavioral anchors (6 domains × 8 abilities) | To author — upstream of the AI evaluator; the other next design task |
| AI evaluator — Eight Direction Abilities feedback | To build from the v2 spec; anchors first, then tune on pilot submissions |
| Direction Record template (one document, eight sections + declaration) | To create |
| Capability profile pages (live, verifiable) | To build |
| Module queue dashboard | To build |
| Foundation content (recorded units) | To record — build directly on the v2 abilities |
| Track A / B module content | Record during pilot cohorts |
| Alumni database + mailer | Live |
| Wednesday workshop | Live — restructure per Marketing tab |
| Pricing — foundation, modules, credential | TO BE DECIDED — open question; keep visible until resolved |
What do I do this weekend?
Weekly rhythm
| When | What |
|---|---|
| Friday (~2 h) | Read the AI evaluator’s weekly summary per track. Pick 2–3 direction trails per active track for weekend critique. Note borderline capstones. Periodically: re-score a small sample across tracks against the anchor exemplars — the moderation routine that keeps a pass meaning the same thing in every domain. |
| Saturday h1–2 | Track C session (rotating module). |
| Saturday h3–4 | Program Slot (funnel session / overflow vivas / open clinic + Skill cold-runs / buffer, by week). |
| Sunday h1–2 | Track A session (high-demand module). |
| Sunday h3–4 | Track B session (high-demand module). |
Session run-sheets
Week 1 — Kickoff (120 min): 0–15 frame the territory in the value ontology · 15–60 live demonstration of directing patterns on a real example (harvested from diagnostics, with permission) · 60–105 participants begin applying to their own work · 105–120 set the week’s application task.
Weeks 2–3 — Critique (120 min): 0–10 recap · 10–90 dissect the sampled trails against the rubric — where framing won, where validation failed · 90–120 harder patterns, Skill-capture review, next task.
Week 4 — Capstone block (120 min): group walkthroughs ~12 min each including viva. AI pre-scores observable evidence on the rubric; the viva, declaration, and supporting evidence confirm what a chat trail cannot show; you confirm or override. Borderline cases go to the Program Slot overflow, never rushed.
Weeks 2–3 — Critique (120 min): 0–10 recap · 10–90 dissect the sampled trails against the rubric — where framing won, where validation failed · 90–120 harder patterns, Skill-capture review, next task.
Week 4 — Capstone block (120 min): group walkthroughs ~12 min each including viva. AI pre-scores observable evidence on the rubric; the viva, declaration, and supporting evidence confirm what a chat trail cannot show; you confirm or override. Borderline cases go to the Program Slot overflow, never rushed.
The Eight Direction Abilities (every submission, every module)
1 Delegatewhat AI should — and shouldn’t — do
2 Framedefine the job and what good means
3 Contextright evidence, background, boundaries
4 Planchoose and sequence the route
5 Steerdiagnose, redirect, know when to stop
6 Validateearn the right to rely
7 Landmake it useful in the real world
8 Capturerepeatable by someone else
An interacting loop, not eight sequential gates — the four beats (Decide · Design · Drive · Deliver) are a teaching mnemonic, never a mandated sequence. The Validate strapline everywhere: “Validate before you rely.” Evidence must be proportionate to complexity, consequence, novelty, and reach — more prompts, tools, context, iterations, or documentation never means a higher score. Full assessor-layer definitions, evidence rules, and ability boundaries: see the v2 specification.
Responsible direction — non-compensable
Six red lines — any one fails the capstone outright, whatever the scores: unauthorized use of confidential, personal, proprietary, or restricted information · fabricated or knowingly unverified material claims, sources, or calculations · consequential authority handed to AI without a human approval point · materially deceptive, discriminatory, manipulative, or unsafe output · deliberate non-disclosure of AI involvement where reasonably required · operational use without necessary permissions, controls, exception handling, ownership, or rollback.
Plus one competence threshold: the participant must be able to explain the output’s limitations, residual risks, and the circumstances in which it should not be used. Not misconduct — but no score elsewhere compensates for it.
Plus one competence threshold: the participant must be able to explain the output’s limitations, residual risks, and the circumstances in which it should not be used. Not misconduct — but no score elsewhere compensates for it.
Viva question bank
- How did you know it was ready?
- Why did you reject the earlier draft? (when drafts were rejected)
- What did you verify before trusting that number?
- What context did the AI need before it became useful?
- What information did you decide AI should not receive — and why?
- Who knows AI was involved in this — and who needed to?
- What breaks if a colleague runs your Skill tomorrow?
- Where did the AI mislead you, and how did you catch it?
- What would you direct differently next time?
- When should this result not be used? (asked in every viva)
Judgment rules: pass = abilities demonstrated on real work; a polished artifact alone proves nothing. One round can demonstrate Steer, ten can demonstrate its absence — iteration count proves nothing by itself, and deciding not to use AI is evidence of Delegate, never a failure. The same AI checking its own work is review, not independent validation. Each module’s capstone has a minimum Land floor (Business Applications: in actual use by at least one other person — no fallback). One structured resubmission after specific rubric feedback. CAID credential: at least one capstone in actual workplace use + at least one Skill cold-run by another person (Wk3 clinic). Profiles list named artifacts, never counts — counts invite inflation. Flag strong CAIDs early: assistant → clinic lead → track owner is the succession plan.
What’s inside?
Learninformation in
Analyzenumbers
Decidechoices
Communicatewords out
Systematizerecurring work
Buildartifacts for others
Six peer domains — no core, no elective, no rank. Four weeks each, ~12–14 hours. Every module: the same Eight Direction Abilities, a real-work capstone that can be failed — with a module-specific minimum bar for real-world use — and named reusable Skills into the capability profile.
1 · AI for Research & Knowledge — turn information overload into trustworthy understanding
For: consultants, strategists, analysts, project leaders, anyone drowning in reading.
Units: framing the question and evidence standard · deep research and source strategy · document and knowledge intelligence · synthesis, comparison, and contradiction · rapid domain onboarding · staying current and material change · validation, provenance, and asset capture.
Skills: Evidence-Backed Research Brief · Document Intelligence Brief (e.g., Long-Document Brief) · Comparison Brief (e.g., Vendor Comparison) · Current-Awareness Digest (e.g., Weekly Industry Digest) · Get-Up-To-Speed Tutor.
Capstone: an evidence-backed brief answering a real, live question for a defined user — claims traced to sources, contradictions surfaced, validation trail visible, reviewed or used by the intended reader, captured as a reusable Skill.
Units: framing the question and evidence standard · deep research and source strategy · document and knowledge intelligence · synthesis, comparison, and contradiction · rapid domain onboarding · staying current and material change · validation, provenance, and asset capture.
Skills: Evidence-Backed Research Brief · Document Intelligence Brief (e.g., Long-Document Brief) · Comparison Brief (e.g., Vendor Comparison) · Current-Awareness Digest (e.g., Weekly Industry Digest) · Get-Up-To-Speed Tutor.
Capstone: an evidence-backed brief answering a real, live question for a defined user — claims traced to sources, contradictions surfaced, validation trail visible, reviewed or used by the intended reader, captured as a reusable Skill.
2 · AI for Analytics — raw data to defensible, reproducible insight that drives action
For: analysts, finance, operations, quality — anyone who turns spreadsheets or business data into decisions.
Units: framing the analytical question and metric · data readiness and provenance · patterns, segments, and exceptions · method selection and assumption testing · result to defensible insight · visualization and the analytical story · validation, reproducibility, and capture.
Skills: Metric & Data Readiness Auditor (e.g., Data Cleaner) · Performance & Variance Analyzer (e.g., Monthly Report Analyzer) · Pattern & Exception Explorer · Forecast & Scenario Support · Insight-to-Action Brief.
Capstone: a real dataset answering a live business question — metrics defined, preparation trail visible, material calculations independently validated, uncertainty stated, reviewed by the intended business user, captured as a reusable Skill.
Units: framing the analytical question and metric · data readiness and provenance · patterns, segments, and exceptions · method selection and assumption testing · result to defensible insight · visualization and the analytical story · validation, reproducibility, and capture.
Skills: Metric & Data Readiness Auditor (e.g., Data Cleaner) · Performance & Variance Analyzer (e.g., Monthly Report Analyzer) · Pattern & Exception Explorer · Forecast & Scenario Support · Insight-to-Action Brief.
Capstone: a real dataset answering a live business question — metrics defined, preparation trail visible, material calculations independently validated, uncertainty stated, reviewed by the intended business user, captured as a reusable Skill.
3 · AI for Problem Solving & Decision Making — the Benchmark signature module
For: managers, leaders, consultants, improvement professionals — the bridge for belt alumni. Improvement methods (DMAIC, A3, issue trees) run through it as a lens, never a toolbox.
Units: the problem or decision and its ownership · framing and decomposition · cause hypotheses and evidence · options and decision criteria · trade-offs under uncertainty · premortems and adversarial challenge (red-teaming your own plan) · decision record, action, and learning.
Skills: Problem Framing & Decomposition Canvas · Root Cause Explorer (hypotheses with evidence, never pronouncements) · Options & Trade-off Evaluator · Premortem & Scenario Challenger · Decision Record & Review.
Capstone: a real problem or decision the participant owns or materially influences, taken to a defensible decision, experiment, recommendation, rejection, or justified deferment — adversarially challenged, considered by the decision owner, captured as a reusable Skill.
Units: the problem or decision and its ownership · framing and decomposition · cause hypotheses and evidence · options and decision criteria · trade-offs under uncertainty · premortems and adversarial challenge (red-teaming your own plan) · decision record, action, and learning.
Skills: Problem Framing & Decomposition Canvas · Root Cause Explorer (hypotheses with evidence, never pronouncements) · Options & Trade-off Evaluator · Premortem & Scenario Challenger · Decision Record & Review.
Capstone: a real problem or decision the participant owns or materially influences, taken to a defensible decision, experiment, recommendation, rejection, or justified deferment — adversarially challenged, considered by the decision owner, captured as a reusable Skill.
4 · AI for Communication & Influence — communication that earns attention, trust, and action
For: managers, consultants, sales, trainers, executives — anyone whose effectiveness depends on how ideas are understood and acted upon.
Units: communication outcome and ethical boundary · audience and stakeholder insight · message architecture and storyline · authentic voice · documents, proposals, and presentations · adaptation across stakeholders, channels, and live conversations · quality, delivery, feedback, and capture.
Skills: Communication Brief & Audience Map · Storyline Architect · Executive Document Builder (e.g., Proposal Writer, Executive Summary Builder) · Presentation Narrative Builder · Stakeholder Message Adapter.
Capstone: a real high-stakes communication plus its influence plan — facts verified, one purposeful adaptation for another audience or channel, delivered or reviewed by the intended recipient, feedback captured, reusable Skill. Ethical influence and disclosure are red-line requirements.
Units: communication outcome and ethical boundary · audience and stakeholder insight · message architecture and storyline · authentic voice · documents, proposals, and presentations · adaptation across stakeholders, channels, and live conversations · quality, delivery, feedback, and capture.
Skills: Communication Brief & Audience Map · Storyline Architect · Executive Document Builder (e.g., Proposal Writer, Executive Summary Builder) · Presentation Narrative Builder · Stakeholder Message Adapter.
Capstone: a real high-stakes communication plus its influence plan — facts verified, one purposeful adaptation for another audience or channel, delivered or reviewed by the intended recipient, feedback captured, reusable Skill. Ethical influence and disclosure are red-line requirements.
5 · AI for Workflow & Productivity Systems — redesign recurring work so the time stays reclaimed
For: knowledge workers, team leaders, operations professionals. The rule: eliminate or simplify first, redesign second, divide human-AI work third, orchestrate only what remains.
Units: finding and framing the right recurring workflow (continues from your report) · current-state mapping and value leakage · the human-AI operating model · reusable routines and light orchestration · handoffs, approvals, exceptions, and sensitive information · personal and team patterns · measurement, governance, and improvement.
Skills: Workflow Mapper & Redesign Canvas · Recurring Status & Report System (e.g., Weekly Status Generator) · Inbox or Request Triage System · Meeting-to-Actions System · Review Preparation System.
Capstone: one recurring workflow redesigned end to end and run in practice — baseline established, exceptions and fallback tested, before/after measured, owner feedback, sustainable ownership, reusable Skill.
Units: finding and framing the right recurring workflow (continues from your report) · current-state mapping and value leakage · the human-AI operating model · reusable routines and light orchestration · handoffs, approvals, exceptions, and sensitive information · personal and team patterns · measurement, governance, and improvement.
Skills: Workflow Mapper & Redesign Canvas · Recurring Status & Report System (e.g., Weekly Status Generator) · Inbox or Request Triage System · Meeting-to-Actions System · Review Preparation System.
Capstone: one recurring workflow redesigned end to end and run in practice — baseline established, exceptions and fallback tested, before/after measured, owner feedback, sustainable ownership, reusable Skill.
6 · AI for Business Applications — working tools without becoming a developer
For: business excellence, operations, CI teams, technically curious managers. Narrower audience — narrower, not higher; the diagnostic routes the right people here, often after Module 5. Assessed on direction judgment, never coding volume or polish — a key ability is knowing not to build.
Units: choosing and safely scoping the right application · users, workflow, and requirements · data, rules, permissions, and interface · AI-assisted building · application patterns and light integration · testing exceptions, permissions, and misuse · deployment, ownership, and handover.
Skills: Small-System Blueprint · Data & Rule Modeler · Form, Tracker & Approval Builder · Dashboard & Reporting App Builder · Test, Deployment & Handover Pack.
Example artifacts: approval tracker, operations dashboard, inventory utility, internal reporting app.
Capstone: one bounded working tool used by at least one other person to complete a real task with reasonable independence — tested, permissioned, documented, owned, handed over. No fallback on this floor; a guided demonstration is not enough.
Units: choosing and safely scoping the right application · users, workflow, and requirements · data, rules, permissions, and interface · AI-assisted building · application patterns and light integration · testing exceptions, permissions, and misuse · deployment, ownership, and handover.
Skills: Small-System Blueprint · Data & Rule Modeler · Form, Tracker & Approval Builder · Dashboard & Reporting App Builder · Test, Deployment & Handover Pack.
Example artifacts: approval tracker, operations dashboard, inventory utility, internal reporting app.
Capstone: one bounded working tool used by at least one other person to complete a real task with reasonable independence — tested, permissioned, documented, owned, handed over. No fallback on this floor; a guided demonstration is not enough.
Typical pathways
Finance manager → Analytics + Problem Solving · Consultant → Research + Communication · Operations manager → Workflow + Business Applications · Trainer → Communication + Research · HR manager → Communication + Workflow. The diagnostic recommends; all six stay open to everyone. Credential rule: Foundation + demonstrated competence in any two domains → CAID, modules listed as endorsements.
