Skip to content
TastePolicy.Open TastePolicy

Judge walkthrough · no login

Margaret's mobility changed. Her taste didn't.

Qloo understands what feels like Margaret. TastePolicy decides whether her AI may act on that understanding. Three short acts show the difference with real Qloo data.

Act 1 · The person

Meet Margaret

She still knows exactly what she enjoys. Her circumstances changed; her taste did not.

  • Miles DavisIntimate live jazz
  • Ella FitzgeraldVoices she grew up with
  • CasablancaClassic cinema
  • Cinema ParadisoClassic cinema
  • EatalyTuscan & Italian dining
  • Step-free entrance · requiredNot checked yet
  • Walk of 10 minutes or less · requiredNot checked yet
  • A quieter room · preferredNot checked yet
  • Within $75 · preferredNot checked yet

Request: “Somewhere for a lovely evening out, nothing too far, and not too loud.” AI permission: suggest, compare and shortlist — nothing more.

Act 2 · Cultural intelligence

Watch Qloo understand what fits her

Qloo resolves Margaret's five cultural anchors to real entities, discovers cross-domain places from those signals and provides candidate-specific cultural evidence. TastePolicy keeps that evidence separate from accessibility, walking time, noise, price and every other real-world requirement.

Generic AI

A plausible suggestion. It can sound personal without grounded cultural evidence.

Qloo

The same choice becomes culturally grounded: this candidate fits Margaret's cross-domain taste, with evidence for why.

TastePolicy

Now ask the question recommendation alone cannot answer: does it work for her real needs, and may the AI go this far with it?

Place photos, maps, street sequences and 360° media are presentation. They never become factual evidence merely because they look relevant.

Act 3 · Agency

Watch TastePolicy know when to stop

  1. Qloo finds a strong cultural match. Margaret still requires a step-free entrance.
  2. If that critical fact is unknown, TastePolicy returns ASKinstead of guessing: “This feels like Margaret, but I need to confirm the entrance.”
  3. Answer the missing fact. The same evaluator checks the same candidate again.
  4. If the requirements are satisfied, the result can become ALLOW — but only as far as Margaret's saved ceiling: SHORTLIST.
  5. If the agent tries to present that option as the final choice, the authority ceiling forces ASK again.
  6. “Why this happened” separates Qloo evidence, place facts, Margaret's requirements, her rules and the final decision.

Needs your answer

Something important is still unknown, so your AI stops and asks instead of guessing.

Go ahead

This fits you and what works today. Your AI may go as far as you allowed.

A known hard violation becomes REJECT. If required evidence cannot be obtained, the result is UNAVAILABLE. Two concise variants show the same contract beyond Margaret: sensory-aware and low-energy.

The thesis

Taste intelligence is not permission.

Qloo answers what fits this person? TastePolicy answers may this agent go this far with that fit, for this person, now?

Try the live experience

Optional deeper track

Inspect the technical proof

The product idea is complete above. Open this section only if you want the underlying authority contracts, developer surfaces and evidence trail.

Show technical evidence

Qloo is the cultural evidence layer

TastePolicy resolves cultural anchors through Qloo, verifies the identities being used, discovers cross-domain candidates and requests candidate-specific affinity and explainability. It does not recreate Qloo's graph or substitute generic LLM guesses for cultural evidence.

Open the Assisted Choice experience · Inspect the engineering route

One deterministic authority contract

Known hard violation → REJECT. Critical fact missing → ASK. Required evidence unavailable → UNAVAILABLE. Requirements satisfied and the requested action within the person's ceiling → ALLOW.

Qloo evidence, place facts, requirements, rules and the final decision remain separately attributable in Cultural Receipts.

Developer and trust surfaces

The same authority contract is exposed through REST, the TypeScript SDK, remote MCP, A2A and OAuth. Deeper controls include versioned policies, bounded leases, drift checks and portable proofs.

Developer platform · Engineering demo · Proof verifier · Trust boundary · Live readiness

Evidence, not adjectives

Hosted, live, controlled, synthetic and blocked evidence are labelled separately. The public repository includes release, acceptance, benchmark and human-pilot records rather than turning unproved integrations or test fixtures into product claims.

Open the public evidence ledger ↗

The execution boundary stays intact

ALLOW grants preference-choice authority only. It never grants payment, booking, messaging, account changes, private-data disclosure or legal authority. Recommendation is not authorization, and verifying a proof does not make a separate execution permissible.