Platform

Turn implicit judgement into explicit, inspectable persona artefacts

Persona-x creates structured AI personas with quantified judgement profiles. Assemble panels where multiple perspectives challenge, validate, and improve every decision.

The problem

Single-perspective AI creates invisible risk.

Blind spots go undetected

A single AI perspective cannot systematically challenge itself. If no one asks about downside risk, it is never raised — and critical gaps go unnoticed until it is too late.

Judgement is invisible

When an AI gives advice, the reasoning posture behind it — how cautious it was, what evidence it required, what it ignored — is not stated or inspectable.

Consistency is accidental

Without a defined profile, behaviour shifts with phrasing, context, and prompt design. Different users get different levels of rigour from the same system.

How it works

Three stages: Create, Store, Run.

01

Create

A guided engine extracts judgement signals through focused questions — decisions, comparisons, and spectrum choices — and builds a complete persona file scored against a six-dimension rubric.

02

Store

Each persona becomes a structured YAML file: human-readable, machine-validatable, version-controlled, and comparable across the entire persona library.

03

Run

Personas are loaded into a panel session. The runtime selects who speaks, orders contributions by intervention frequency, and translates rubric scores into AI behavioural rules.

The rubric

Six dimensions that define how a persona thinks — scored 1–10, directly comparable across every persona.

Risk appetite

From conservative and downside-focused, through balanced, to bold and signal-trusting.

Evidence threshold

How much documented evidence is required before the persona will support a decision.

Tolerance for ambiguity

Comfort operating without complete information or fully resolved questions.

Intervention frequency

How often the persona speaks up, and how it picks the moments with most impact.

Escalation bias

Whether concerns are handled locally or escalated on material risk.

Delivery vs rigour

The balance between shipping quickly and demanding thorough validation first.

Use cases

Wherever multiple perspectives improve outcomes.

Board decision review

Stress-test board-level proposals against risk, compliance, and commercial personas before they reach the boardroom.

AI system review panel

Evaluate AI deployments with personas covering safety, bias, performance, and governance.

Legal case strategy

Test strategies against personas representing opposing counsel, judicial reasoning, and client risk tolerance.

Clinical decision support

Multi-perspective review of treatment options across clinical evidence, patient preference, and resource constraints.

Hiring decision panel

Evaluate candidates from cultural fit, technical depth, and team-dynamics perspectives.

Career transition adviser

Weigh career moves across financial prudence, personal fulfilment, and market reality.