> For the complete documentation index, see [llms.txt](https://docs.linuxfoundation.org/lfx/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.linuxfoundation.org/lfx/project-control-center/v2-latest-version/collaborations/meetings/verify-meeting-participants/auto-reconcile-user-guide.md). # Auto-Reconcile User Guide ## What is Auto-Reconcile? Auto-Reconcile is an AI-powered feature that helps you quickly match **unverified meeting attendees** to **meeting invitees** after a Zoom meeting has ended. Instead of manually matching each attendee one by one, the AI analyzes names, emails, and other signals to suggest matches for you. *** ## Common cause of "Unknown" attendees: joining by Zoom ID instead of the Join link The most common reason a properly invited attendee shows up as **Unknown**, even when they're a confirmed committee member: they joined the meeting by entering the **Zoom meeting ID and passcode directly** (in the Zoom app or website) instead of using the **Join** link/button on the meeting page in PCC. Joining this way leaves no join-page log for that attendee, so there's nothing for reconciliation to match them against — regardless of whether they were correctly invited. This is separate from, and more common than, the case where a name is simply missing from the LFX user service. If you see a confirmed invitee marked Unknown, check first whether they may have joined via the raw Zoom ID/passcode rather than the Join link, and correct it manually in the **Unverified Attendees** table if so. ## When to use it Use Auto-Reconcile when a past meeting has a long list of unverified attendees that need to be matched with the meeting's invitee list — for example, after a large community meeting or governing board call. *** ## How it works When you click **Auto-Reconcile**, the system sends the list of unverified attendees and the list of invitees (who did not join) to an AI service. The AI returns three categories of results: | Category | Confidence | What happens | | ---------------- | ------------------------------------------------------ | ------------------------------------------ | | **Auto-matched** | **High** — matched via LF SSO or exact email | Applied automatically, no review needed | | **Needs review** | **Medium / Low** — matched by name or email similarity | Shown to you for review before applying | | **Unknown** | No match found | Automatically marked as "Unknown Attendee" | *** ## Step-by-step instructions ### 1. Open the Unverified Attendees table Navigate to a past meeting that has unverified attendees. You will see the Unverified Attendees table. ### 2. Click the Auto-Reconcile button At the top of the table, you will see an info banner with an **Auto-Reconcile** button (marked with a magic-wand icon). > ⚠️ **Important:** The Auto-Reconcile action **cannot be stopped once started**. All transactions it creates are **permanent** and can only be corrected by contacting support. Make sure you are ready before clicking. The button will change to **"Reconciling..."** with a spinning icon while the AI is processing. This usually takes a few seconds. ### 3. Review the summary banner When Auto-Reconcile finishes, a summary appears next to the button: * **X need review** (yellow magnifying glass icon) — matches that need your attention * **X unknown** (grey question-mark icon) — attendees with no match found Hover over each badge for a tooltip showing the exact counts and percentages. ### 4. Review medium/low confidence matches If any matches need review, a yellow panel appears below the banner showing each suggested match: * **Attendee name** → **Suggested invitee name** * A **confidence badge** (Medium or Low) * An **info icon** — hover to see the AI's reasoning for the suggestion * **Signal tags** (blue) showing what signals the AI used (e.g. similar name, email domain match) For each match, you have two options: * **Accept** (green button) — Confirms the match and applies it. The attendee is marked as verified. * **Reject** (white button with red text) — Discards the suggestion. The attendee remains unverified, and you can match them manually. If there are many matches, use the pagination dots at the bottom to page through them. ### 5. Handle "Unknown" attendees Attendees in the **unknown** category are automatically marked as "Unknown Attendee". No action is required from you for these. If you later determine who they were, you can still edit them manually from the table. ### 6. Auto-matched entries You don't need to do anything for these — they are already applied. They will disappear from the Unverified Attendees table and appear in the verified attendee list. *** ## Tips * **Review before accepting.** The Medium/Low confidence suggestions are the AI's best guess — always verify the reasoning (hover the info icon) before clicking Accept. * **Rejecting is safe.** If you reject a suggestion, nothing is saved. You can still match the attendee manually using the dropdown in the Unverified Attendees table. * **Check the signals.** The blue signal tags (e.g. "email domain match", "name similarity") give you a quick idea of why the AI suggested this match. * **Use pagination.** When there are many review matches, use the dots at the bottom of the review panel to navigate through pages. *** ## Frequently Asked Questions ### Can I undo an auto-matched attendee? Auto-matched attendees (high confidence) are applied immediately. If an auto-match was incorrect, you can correct it by editing the attendee manually or by contacting support. ### Can I undo an accepted review match? Once you click **Accept** on a review match, the match is applied. To undo it, edit the attendee manually from the Unverified Attendees table or contact support. ### What if the AI service fails? If the AI service is unreachable, you will see an error message: *"Could not reach the AI service. Please try again or contact support."* Try clicking Auto-Reconcile again, or fall back to manual matching. ### Does Auto-Reconcile work for future meetings? Auto-Reconcile works on a single past meeting at a time. For each past meeting with unverified attendees, you need to click Auto-Reconcile separately. ### What data does the AI see? The AI receives the list of **attendees** (who joined the meeting) and **invitees** (who were invited but did not join as verified attendees), along with names and emails. It uses these to suggest matches. No other personal data is sent. *** ## Need help? If Auto-Reconcile produces unexpected results or you need to correct a transaction after it was applied, please contact LFX support. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://docs.linuxfoundation.org/lfx/project-control-center/v2-latest-version/collaborations/meetings/verify-meeting-participants/auto-reconcile-user-guide.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.