WorkPortfolioThe ThesisAboutContact中文
All work
In Pilot · Q2 2026 · Consent layer

Aura

Trust & consent for being learned by AI

Before AI can learn the real world, a human has to say yes. Aura is the layer that earns that yes — capturing what people commit to, what they consent to, and what the system is allowed to remember.

AThe consent mechanism of the Field Data Firewall
Role in firewall
Consent & trust mechanism
Status
Piloting · WEF Global Shapers Taipei
Surface
WhatsApp Cloud API
Stack
Claude Sonnet · Supabase

The problem.

The bottleneck to AI entering real life isn't capability — it's consent. People won't let a system into their work, space, and relationships unless they trust what it captures and what it does with it.

Communities run on commitments made in chat and quietly dropped. The information exists; it's trapped in an unstructured stream nobody can query — and nobody has agreed to anyone learning from.

The approach.

Aura sits inside WhatsApp, where the community already lives — no new app to adopt. It builds a trust, permission, and memory layer in the open.

  • Captures stated commitments from natural conversation
  • Makes consent explicit — what is remembered, and what is not
  • Runs recruiting, onboarding, and project pulses as light flows
  • Turns chat history into a queryable, permissioned operations memory

It's the human-facing front door of the firewall: the place where people decide, on their own terms, what AI is allowed to learn.

Where it stands.

Aura is piloting with the World Economic Forum's Global Shapers Taipei Hub — a real community with real coordination load, not a demo.

Q2 26
Active pilot
WA
Native surface
Consent
By design
Next project
litell →
Work with PlexMesh