Rosie is an AI assistant for a professional network that doesn't just answer questions — it produces answers you can stand behind: grounded in your network's own curated knowledge, traceable to their sources, and — when it matters — confirmed by the real experts behind them.

This page explains how that works, in two parts: the lifecycle of an answer, and the ecosystem that makes the network stronger the more it's used.

Part 1 — The lifecycle of an answer

The lifecycle of an answer — and where Rosie goes beyond answer-and-cite.

1 & 2 — Seeding the knowledge

A member of the association's governing body starts the network by contributing curated, authoritative content and inviting members in. Members then add their own curated content. Every contribution carries Document Metadata — the contributor and the source attributions — so provenance is built in from the very first document. The governing body can also invite supply-chain organizations to contribute, broadening the knowledge base while keeping it governed.

3 — Asking Rosie

A member puts a question to the network's knowledge, and Rosie returns an answer with its sources cited. This is the part that looks familiar: NotebookLM — which many teams already use — along with Microsoft Copilot, ChatGPT, Claude, and Gemini, all answer a question and cite where the text came from. Citing sources is no longer what sets a tool apart.

The Rosie layer — what rides on every answer

This is where Rosie does what the others don't. Every answer arrives with:

• a confidence dashboard showing how strongly each source supports the answer;
• content provenance and source attribution — a clear trail back to the contributed material;
• Recognition of Contribution (RoC) credits for the members whose content was used;
• IP rights (IPR) licensing revenue, with optional derivative licensing, so contribution is compensated, not just acknowledged.

The result is an answer that is accountable, traceable, and rewards the people behind it — none of which the general-purpose tools provide.

4 — Verification (optional)

When an answer matters enough to warrant more than provenance, members bring real experts into the loop, in up to three escalating levels — each optional, each building on the one before:

• Level 1 — Semantics. Members invite the original source contributors to confirm their contribution is represented accurately. These services run on peer-to-peer Agreements that set who verifies, the scope, and the price — which is what makes this a genuine two-sided market.
• Level 2 — Holistic. Once Level 1 is complete, a governing-body member can be invited to confirm the response in its entirety.
• Level 3 — Final. An outside party — such as legal counsel — can be invited for a final layer of assurance.

Members can also author their own verification procedures, defining how a given kind of question should be checked — but that's a deeper capability than this overview needs.

Part 2 — The ecosystem that makes it stronger

The reward loop, discoverability, and how the network compounds as peers join.

The reward loop

Recognition of Contribution credits and IPR licensing revenue flow back to the people whose content and verification made an answer trustworthy. Contributing and verifying are rewarded — which draws more of both — which makes answers more trustworthy — which drives more use — which generates more reward. The network is a self-reinforcing economy, not a static database.

Discoverability is the hinge

Because every answer cites the contributions behind it, every cited contributor becomes a discoverable expert: named, and visibly tied to a real answer. This turns the membership into a living, competitive directory of experts — ranked, in effect, by whose work actually answers the questions members are asking.

Local discoverability — value on day one

A member asking about, say, EU origins and trade receives an answer citing several of their own organization's experts — and in doing so discovers viable expert alternatives they didn't know were available. This works within a single organization's membership; no wider network is required for it to matter.

Network discoverability — how the network compounds

When that same answer also draws on another peer organization's contributor, a new expert relationship is created that couldn't have existed before. Each peer that joins makes its contributors discoverable to every other member — widening the expert market for all of them. Adding peers costs Rosie nothing architecturally; the network simply grows more valuable as it grows larger.