Building an AI Tarot Decision Engine: Turning Seven Dimensions into Structured Reports

David Webb — Tue, 15 Sep 2026 12:39:58 +0000

Some weeks I ship everything on my list. Other weeks I stare at a half-written email for two hours and call it a day. I used to blame focus, sleep, or motivation — until I started treating my decisions like data instead of feelings.

That habit turned into a product: an AI-assisted decision report engine that frames a structured analysis as a tarot-style reading. This post is about the model underneath — no mysticism required — and what building it taught me.

The observation

For years, whenever I faced a decision I was avoiding, I'd pull tarot cards. Not for prediction — but because a spread forces you to look at the situation from angles you'd rather skip. Positions like "what you fear" or "what you refuse to see" are basically canned journal prompts.

I noticed I kept asking the same seven questions. So I wrote them down, scored them, and mapped them onto continuous dimensions.

The seven dimensions

Every decision report reduces the situation to seven indices on a 0–100 scale (in practice calibrated to an observed 12–88 band, because humans rarely live at the extremes):

Dim	Name	What it measures
R	Risk appetite	tolerance for irreversible downside
I	Intuitive synthesis	how much you trust pattern-recognition vs. explicit reasoning
P	Temporal patience	willingness to wait for compounding
S	Social dynamic	how much alignment/consensus you need
A	Action bias	commit-fast vs. deliberate
Rf	Reflection depth	tendency to re-analyze
St	Stability	preference for known structures over flux

A quiz (about 3 minutes) scores the question, not the person. The output is a seven-dimension radar plus the two or three tensions hiding inside it — e.g. high reflection + low action under load reads as "decision fatigue wearing a costume."

Archetypes as geometry

Ten recurring vectors fell out of the calibration data — each is just a shape across the seven axes:

Archetype	R	I	P	S	A	Rf	St	Failure mode
Alchemist	66	22	58	30	64	44	34	solitary conviction without alignment
Sage	40	78	74	52	20	72	56	over-analysis; missed asymmetric upside
Challenger	76	40	28	44	84	30	32	hyperactive interference; burn rate
Builder	44	62	74	52	50	58	76	rigidity past optimal pivot points
Creator	60	30	30	42	76	30	28	draft overload; consolidation lag
Opportunist	74	50	34	46	72	34	30	shallow commitment; option chasing
Architect	38	78	76	50	58	74	76	inelasticity in fast-changing terrain
Leader	56	56	64	82	72	52	64	high consensus load; diluted velocity
Maverick	76	40	30	22	78	30	26	key-person isolation
Competitor	70	50	40	58	78	40	48	scoreboard fixation

These aren't personality verdicts — they're operating modes. The same person flips vectors depending on the domain (work vs. relationships vs. money).

Engineering notes

Quiz-gated queue. Free beta is 3 readings/day, gated by the quiz so the queue stays human-scale and bots can't burn the LLM budget.
The LLM does framing, not fortune-telling. It renders tensions and next steps from the scored dimensions; the tarot layer is a presentation skin over a structured report.
Honest statistics. The birth-timestamp mapping that seeds the archetype guess is a weak statistical correlation. The product says so, in the UI and in the copy. Users trust it more, not less, for that.
Naming is the intervention. The most common feedback: "I froze less once the pattern had a name." A labeled failure mode is fixable; unnamed anxiety isn't.

What I'd do differently

Ship the radar before the tarot skin. The mystical frame is great for reach, but the retention comes from the structured report.
Score the question first, person second. Question-scoring generalizes; profiling feels invasive.
Ship with the disclosure. Every A/B test said honesty converts better than mystique.

If you're stuck in a decision loop, try tracking one dimension — risk or action — for a week. Naming the pattern is half the fix.

I'm building this in public — the engine is in free beta (tangoera.com) and the longer build log lives on Indie Hackers. Happy to answer questions about the calibration data or the quiz design.

Building an AI Tarot Decision Engine: Why I Used Tarot as a Framing Layer (and What the 7-Dimension Radar Actually Models)

David Webb — Tue, 15 Sep 2026 12:39:10 +0000

A few months ago I caught myself doing the thing every solo founder does: re-reading the same pros-and-cons list for the fourth time, hoping the answer would change. I wasn't lacking information. I was stuck in a loop — re-analyzing instead of committing.

That loop is what I ended up building a product around. TangoEra is an AI-powered decision report tool, currently in a free beta. This post is the engineering story: why tarot, what the engine actually computes, and the mistakes I made along the way.

Why tarot is a framing layer, not a prediction layer

The obvious objection: "AI tarot? So it predicts the future?" No — and this distinction is the whole design.

Tarot survived for centuries not because it predicts outcomes, but because a card spread is a forced perspective machine. The Fool, the Tower, the Two of Swords — each card is a lens that drags your thinking onto an axis you were avoiding. When you draw "restraint" on a day you're fired up to quit your job, the value isn't mystical. It's a structured interruption of your default narrative.

So the product framing is: the AI generates an archetype profile and a decision report, and the tarot visual layer is the interface that makes people slow down and actually read the analysis instead of skimming bullet points. The cards are UI, not oracle. Every report carries the honest disclaimer: this is a self-reflection and decision-support tool based on statistical weak correlation, not prediction, not advice.

I think of it as putting a decision matrix inside a story people will finish.

The engine: archetype vectors, not a black box LLM

Under the hood there are two layers.

Layer 1 — the calibration engine. We model each user's decision style as a vector over seven continuous dimensions: risk appetite, intuitive synthesis, temporal patience, social dynamic, action bias, reflection depth, and stability. Scores live on a 0–100 index, but real profiles only occupy an observed 12–88 band, which matters for normalization.

We maintain 10 prototype archetype vectors — the Alchemist, the Sage, the Challenger, the Builder, the Architect, the Maverick, and so on. Each archetype is a distinct geometric shape across the seven dimensions, with an operational strength and a primary failure mode. The Sage scores high on reflection and patience, low on action bias — brilliant at verification, prone to missing asymmetric upside through over-analysis. The Challenger is the mirror image. When a user's quiz answers come in, we project them onto this prototype space (nearest-shape matching with soft membership, not hard classification) rather than asking an LLM to freestyle a personality type. That keeps output stable, testable, and cheap to regression-test.

Layer 2 — the report radar. The seven dimensions the user sees in their report are translated into decision-facing axes: clarity, risk appetite, timing, relationships, resources, blind spots, and next move. This is where the tarot cards bind to computed values — a card is selected because a dimension score (plus its distance from the archetype's own baseline) triggers it, not at random.

The tension rendering is the part I'm proudest of. A report that says "you're high risk, high reflection" is useless. A report that says "your reflection score is 38 points above your action score — that gap is your frozen-week pattern, and here is the specific next step sized to close it" is a tool. Reports render two sections: tensions (pairs of dimensions that are far apart and therefore in internal conflict) and next steps (one concrete, low-commitment action per tension). One next step, not five. Choice overload is the disease; a report with five action items is just more of it.

This borrows directly from the decision-matrix tradition (Stuart Pugh's concept-selection matrices): score dimensions, weight by importance, let arithmetic surface structure instead of opinions. Our quiz is essentially a weighted scoring instrument wearing a tarot costume.

The 3-minute quiz as a queue gate

The quiz exists for two reasons. First, calibration: we need ~20 weighted answers to place you in the seven-dimensional space with enough confidence that the radar isn't noise. Second — and this is a product decision, not a technical one — it gates the daily report queue.

We generate a limited number of full AI-written reports per day, with a daily 3 seats for the free beta. The quiz is the gate: it filters out drive-by traffic and makes sure every generated report gets a human who actually engaged with their inputs. As a solo dev, this isn't artificial scarcity — it's capacity management. Each report is reviewed for template drift, and human-scale volume is the only way I can catch the LLM quietly going off-script. If you've ever run a generative feature unattended, you know the failure mode: output slowly drifts and nobody notices for weeks.

Honest limitations

The birth-timestamp-to-dimension mapping is a weak statistical correlation, not science. It's a vocabulary generator, not a verdict. I'm upfront about this in-product.
Quiz self-reporting has all the usual biases. People answer who they want to be on Monday morning.
The report is a mirror. If a user wants certainty, no tool sells that honestly.

Solo-dev lessons

Constrain the LLM with geometry. Let the model write prose inside a computed skeleton (archetype vectors, tension pairs, threshold rules), never invent the skeleton. Testability went from "vibes" to unit tests on vectors.
Human-scale queues are a feature. A daily cap forces review loops, which catch drift, which builds trust.
Own the weird framing. Tarot gets attention that "decision support SaaS" never would, and the card constraint genuinely improves the writing — forced metaphors beat blank pages.

If you're building something similar, steal the pattern: compute hard, narrate softly, and always label the correlation honestly.

Try the free beta: https://tangoera.com/?utm_source=devto

DEV Community: David Webb