The Insight Navigator™ — usage protocol
What this tool is
The Insight Navigator is a research methodology and insights advisor. It helps you design rigorous research, generate genuine insight from data, and communicate findings with confidence. Its advice is grounded in a locked knowledge base authored by Kellie Lyons of Perspicacious, built on the Powerful Insight™ framework.
It is an AI system. It is a mentor, not an oracle, and not an approval.
What you may put in
- Publicly available or de-identified content.
- Draft survey instruments and discussion guides.
- Methodology questions, analysis approaches, and dummy or aggregate results.
- Your own thinking, in draft, however rough.
What you may not put in
- Personal information — names, contact details, or anything else that identifies a real person.
- Health information, or sensitive information: racial or ethnic origin, political opinions, religious or philosophical beliefs, trade union or association membership, sexual orientation, criminal record.
- Raw response data about real people. De-identify it first, or ask the Navigator to work at a general, illustrative level.
- Confidential Council information that has not been cleared for release.
The Navigator will refuse content that appears to breach these rules, explain why in a line, and offer to continue once the material is de-identified. That refusal is a backstop, not a filter to rely on — the judgement is yours before you paste.
Attribution — the line that travels
Where the Navigator drafts or substantially rewrites text you carry into a Council deliverable, this line goes with it:
This material was developed with the assistance of an AI research tool (The Insight Navigator™) and has been reviewed by [name].
Replace [name] with the person who reviewed the material before use. Where AI-assisted analysis contributes substantively to a report, the methods statement should say so — the Methods-Statement Checklist in the knowledge base carries the standard wording. This satisfies clause 5.2.6 of the Glen Eira AI Policy.
Where the tool stops
- It is not a legal or privacy adviser. Questions about privacy law, data protection obligations, or what the AI Policy permits go to your privacy or governance contact at Council.
- It is not an ethics approval. Where a matter needs formal ethics review, the Navigator will say so and tell you who to ask.
- It does not run statistics. It advises on what analysis to conduct and how to read the result.
- It does not decide. It structures the thinking; you own the decision, and accountability rests with you (clause 5.2.8).
- Beyond its knowledge base, it refers. Questions outside scope come back to Kellie Lyons and the Perspicacious team.
What is recorded
- Your conversations are not stored on our servers. They live in your browser tab until you clear them, and in any transcript you download. Closing the tab or starting a new conversation ends them.
- We log metadata only — who signed in, when, how many messages, and token usage. Never message content.
- We hold the participant list, your acknowledgement of the first-use notice, and those access logs — for the licence term plus the Standard Terms' retention period, then destroyed.
- Your inputs go to the AI provider to generate a reply. That flow is the one documented in the privacy impact assessment, and is why the rules above exist.
Your access
- Access is named, revocable, and time-limited. There is no sign-up and no sharing.
- Do not forward your sign-in link. It is yours, and it is logged as yours.
- If something looks wrong — a refusal that seems mistaken, advice that seems off, or a reply that quotes something you did not provide — stop and tell Perspicacious.
The short version
Nothing about a real person goes in. Anything that comes out and gets used carries the attribution line. Anything about what is permitted goes to a human at Council. You remain the researcher; the Navigator charts the course.