AINA
Porcelain v2 · surface audit

Every screen that existed in the v1 system, re-skinned — plus the three entry surfaces that were missing.

Marketing
Entry
Assessment
Product · Luminous OS
Console
Locked materials
Cinzel titling, uppercase
Teal #1C6B65 — the one action
Gold #B88E4C — line & achievement
Flat. Hairlines only, no shadow.
{{ klass }}

{{ title }}

What changed

{{ note }}

AINA
CurriculumPracticeFor teamsPricing
Sign in Start assessment
Learn · Grow · Command

Amplify your
intelligence

A driving school for working with AI. You learn judgement that survives the next model release — not the buttons of one tool.

Start your assessment See the curriculum
Model-agnosticClaude, GPT, Gemini
Role-personalized30 lessons, your work
ScoredRubric on every attempt
AI readiness 6 MIN
68 /100
Strong foundation
Framing16
Verification11
Workflow8
Recommended next Read like a senior reviewer
Learn in context Short lessons built on the work you already do.
Practice hands-on Attempt, get scored, retry. In one thread.
Follow your path Sequenced from your role and readiness band.
Track real progress One score, moving, with its own history.
Find out where you stand

Six minutes, twelve questions, one honest number — and the path that follows from it.

Start assessment
AINA
Learn the thinking,
not the tool

Your path, your score and your attempt history are waiting where you left them.

Model-agnostic · role-personalized
Sign in

New here? Take the assessment first — it builds your path.

Continue
or
Continue with Google Continue with Microsoft

By continuing you agree to the terms and privacy notice.
AINA
Step 02 / 04 Save & exit
Personalize your path

What do you do all day?

Every lesson example, rubric and practice brief is rewritten for the role you pick. You can change this later.

A Marketing & contentBriefs, campaigns, voice, research
B Product & designSpecs, research synthesis, reviews
C OperationsProcess, vendors, reporting
D Finance & analysisModels, variance, board packs
E People & enablementHiring, training, comms
F Something elseTell us in one line
Saved automatically Back Continue
AINA
AI readiness assessment 05 / 12
Section 02 · Working with AI

The output is wrong. What do you do first?

A Rewrite the prompt and run it again.
B Work out which part is wrong before touching the prompt.
C Switch to a different model.
D Fix it by hand and move on.
E Add more context and hope.
Answer saved · 09:12 Previous Next question
AINA
Your readiness report Download PDF
AI readiness
68 /100
Strong foundation
Likely range 62 – 74 Twelve answers is a thin sample. The band narrows as you complete lessons.
Where you're ahead
You frame requests well and you already break big asks into steps. That's the half most people never reach.
Where it costs you
You accept plausible output without a check. Verification is the cheapest score you can move.
Four dimensions
Framing16/20
Verification11/20
Workflow design8/20
Judgement13/20
Your path starts here
1.5Read like a senior reviewer8 MIN
1.8Know what it can't know7 MIN
2.4Validate against the standard9 MIN
Start lesson 1.5 or see the whole path
Scored 6 Aug 2026 · 12 answers · v2 rubric
AINA
Dashboard My path Library Practice Progress
A Alex RiveraPro · 68
Thursday, 6 August

Good morning, Alex

Search lessons⌘K
Continue · level 1 3 OF 5 STEPS
Read like a
senior reviewer
Continue lesson Last attempt scored 14/20
Lessons done 18 ▲ 3 this week
Readiness 68 ▲ 4 since June
Milestone Level 1 clear Earned 4 Aug
Up next
1.8Know what it can't know7 MIN
2.4Validate against the standard9 MIN
2.9The last 20 percent6 MIN
Your posture Emerging LIKELY 62 – 74
Framing
Verifying
Workflow
Judgement
Where this goes Three verification lessons would move you into Practising — the band where AI stops costing you rework.
Updated 09:12 · from 6 attempts
Marketing & content · 30 lessons

My path

Completed18 / 30 Re-take assessment
Level 1 · Foundation 10 / 10 ★ Cleared
1.1Structure a better AI request17/20
1.2Break the monolith15/20
1.5Read like a senior reviewerIn progress
Level 2 · Applied 6 / 10
2.3Brief, don't dump16/20
2.4Validate against the standard9 MIN
2.9The last 20 percent6 MIN
Level 3 · Systems Unlocks at readiness 70
Ten lessons on relays, gates and partnership design — where one person's workflow becomes the team's.
30 lessons · 3 levels

Library

Search all lessons
All Foundation Applied Systems For my role Under 8 min Not started
Foundation · 1.1 Structure a better AI request The four parts every request needs before you press send. DONE · 17/206 MIN
Foundation · 1.5 · current Read like a senior reviewer Reviewing output for what's missing, not what's wrong. 3 OF 5 STEPS8 MIN
Foundation · 1.8 Know what it can't know Where the model's confidence outruns its evidence. NOT STARTED7 MIN
Applied · 2.3 Brief, don't dump How much context is enough, and what to leave out. DONE · 16/209 MIN
Applied · 2.6 Voice is what you never say Teaching a model your register without a style guide. NOT STARTED8 MIN
Systems · 3.1 · locked Architect the relay Handing work between people and models without loss. UNLOCKS AT 7011 MIN
AINA
Level 1 · Lesson 1.5 Read like a senior reviewer
What reviewing is The missing-thing pass Your attempt Compare Apply at work
Up next 1.8 Know what it can't know
Step 03 of 05 · your attempt Read like a senior reviewer
14/20
AINA09:07

A senior reviewer doesn't hunt for errors. They ask what a competent version would contain, then notice what isn't there. Take this summary and tell me the three things it's missing — not the three things it gets wrong.

Q3 churn rose to 4.1%. Enterprise held steady.
Recommend increasing onboarding touchpoints.
No segment split on the 4.1%. No baseline — is 4.1% up from 3.2 or 4.0? And the recommendation isn't tied to a cause.
A
14 /20 Solid reviewer instinct ▲ 3
Found omissions6/7
Named the cause gap5/8
Asked for a baseline3/5
Where to push: you spotted the missing baseline but phrased it as a question. A reviewer states the standard — "churn needs a prior-quarter comparison" — so the writer knows what to add.
Try again Hint 1 Hint 2 Skip — apply at work
Write your second pass… Send
Scored against the lesson rubric · not a chat model's opinion
Your trajectory This lesson
Attempts
ATTEMPT 111/20 ATTEMPT 214/20
Coverage Omission pass Cause gap Stating the standard
Glossary Omission pass — reading for what a competent version would contain. Baseline — the prior figure that makes a number mean something.
Settings
Account AI models Learning Notifications Team Billing

AI models

AINA teaches judgement, not one vendor's interface. Pick the model your practice attempts run against — the lessons and rubrics don't change.

ClaudeSonnet 4.5 · default for practice Active
GPTBring your own key
GeminiBring your own key
My workplace modelAzure, Bedrock or an internal endpoint Configure
Practice behaviour
Show the rubric before I attemptOff means you're scored blind, which is closer to real work.
Use my own work as practice materialPaste a real brief and the lesson adapts to it.
Weekly readiness digestOne email, Monday, with your score movement.
PlanPro · annualRenews 12 Feb 2027 $240/yr Manage
Northwind Group · 42 seats

Team console

Export CSV Invite members
Seats used 42 / 50
Median readiness 64 ▲ 6 since June
Active this week 31 ▼ 2 vs last week
Milestones earned 7 ★ 3 cleared Level 2
Member Role path Readiness Lessons Last active
Alex RiveraMarketing68 ▲418/30TODAY
Priya MenonProduct74 ▲724/30TODAY
Daniel OkoyeFinance61 ▲214/302 DAYS
Sarah MitchellOperations58 ▼111/309 DAYS
Tom HendricksEnablement70 ▲521/30YESTERDAY
Mei ChenProduct—0/30NOT STARTED
As of 6 Aug 2026 · 09:12 UTC · from assessment scores + lesson events. Readiness for members with fewer than 3 attempts is shown as a band, not a point.