Pathway · 02 · For participants

Five exchanges with the AI assistant, free.

A short, voluntary sample of the Global Narrative Atlas platform. Five reflective exchanges with the platform's AI guide, in one sitting. Your transcript is yours to download as a Markdown file at the end.

§ 01 What you'll do

One sitting. Five exchanges. About ten to twenty minutes.

  • You sign in with your email — a one-time link in your inbox; no password to manage.
  • You see a brief consent text covering what data is stored, how, and for how long. You may decline at any moment.
  • The platform's AI assistant opens with a canonical narrative-identity question. You reply at whatever length feels right.
  • The guide responds and asks one follow-up — drawn from the canonical protocol or probing what you just said.
  • Five such exchanges, then the input closes. Your conversation stays visible, and a download button activates.
§ 02 What you'll keep

Your transcript, yours to take.

When the five exchanges are complete, an Export transcript button activates in the room sidebar. You download a single Markdown file containing the full conversation, formatted for easy reading or further annotation.

Sample export filename

gna-open-engagement-anonymous_otter-20260513.md
§ 03 What this is not

A research prototype — not the full longitudinal study.

Not therapy. The platform does not diagnose, treat, or claim wellbeing outcomes. If you're in distress, please use the crisis resources.

Not a chatbot. No character roleplay, no engagement optimisation. The dialogue serves a research methodology.

Not the full study. Five exchanges is a taster. The formal longitudinal study (recruited participants, full 12-question protocol, trajectory reports over months) is a separate path, available to qualifying researchers' cohorts.

Not anonymous to the platform. Your email is linked to your account. Inside the database, the rest of your data uses a randomly generated pseudonym. The platform never sells data, never trains models on it, and never shares it with third parties.

§ 04 Privacy & ownership

Recorded only with your consent. Yours to delete.

Considered storage

Your dialogue is cloud-hosted on Supabase and encrypted at rest, recorded only with your explicit consent at enrolment. The DOL analytical pipeline is open-source and may be inspected on GitHub.

Full data ownership

You retain the right to review, withdraw, or delete any contribution at any time. Deletion cascades — your transcripts, scores, and derived artefacts are removed.

No third-party use

No advertising. No model training on your data. No third-party access. The platform is non-profit research infrastructure, not a product.

Deterministic extraction

The lexicons, scripts, and statistical procedures applied to dialogue are publicly documented. Lexicon-based extraction is deterministic: identical inputs yield identical outputs. Reproducibility in the wider scientific sense — that independent researchers obtain comparable results on independent data, and that the constructs measured correspond to what they purport to measure — is the subject of ongoing work and is not yet established.

§ 05 Governance

Oversight, kept non-extractive.

The platform operates under a governance architecture intended to prevent extractive use of participant narrative. Several elements are currently in formation; the framing here is what the platform commits to, not all of which is fully staffed yet.

  • Technical Council

    Reviews the analytical pipeline, safeguarding events, generated reports, and audit logs. Currently in formation.

  • Cultural Council

    Reviews the cultural fit of canonical questions, lexicons, and report register per language and population. English-only at present.

  • Annual Ethics Ledger

    A published transparency document covering data flows, retention, safeguarding events, and council recommendations.

§ 06 For researchers

Running a longitudinal study? The full platform is a separate path.

The Open Engagement experience above is a five-exchange sample. The full longitudinal platform — your own canonical questions, your own consent text, your own AI-guide behaviour, your own report generation, complete data isolation between studies — is built on the same architecture and is offered as bespoke infrastructure to qualifying research programmes.

Five exchanges. Ten to twenty minutes. Your transcript, your choice.

Create your account →