> ## Documentation Index
> Fetch the complete documentation index at: https://help.terriqon.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Review, Approve, and Download AI-Generated Reports

> Review AI-generated report drafts, confirm PII stripping, then approve or reject. Every report requires a named manager sign-off before download.

Terriqon's AI can turn weeks of field data into a structured, funder-ready report draft in under 15 minutes — but no draft ever leaves your organisation without a human sign-off. As an Organization Admin or Organization Manager, you are the final checkpoint. You review the draft, make edits, verify that PII has been stripped, and either approve the report for distribution or reject it and send it back to draft. This guide explains the full approval workflow from generating a report to downloading the finished document.

## Who Can Approve Reports

Report approval is restricted to users with the appropriate role:

| Role                     | Can Generate Reports | Can Review & Edit | Can Approve or Reject |
| ------------------------ | -------------------- | ----------------- | --------------------- |
| **Organization Admin**   | Yes                  | Yes               | Yes                   |
| **Organization Manager** | Yes                  | Yes               | Yes                   |
| **Field Officer**        | No                   | No                | No                    |

Field Officers can submit data that feeds into reports, but they cannot access the report generation or approval workflow. This separation ensures that sensitive programme narratives and funder communications are only handled by authorised team members.

## Generating a Report

<Steps>
  <Step title="Open the Reports tab">
    Navigate to your project from the **Projects** sidebar and click the **Reports** tab at the top of the project page.
  </Step>

  <Step title="Click Generate Report">
    Click the **Generate Report** button. A setup panel slides in from the right.
  </Step>

  <Step title="Select a report type">
    Choose from one of four report types:

    <CardGroup cols={2}>
      <Card title="Project Report" icon="folder-open">
        A comprehensive narrative covering activities, outputs, and outcomes for a single project. Ideal for internal reviews and project-level funder updates.
      </Card>

      <Card title="Ops Report" icon="gears">
        An operational summary focused on field team performance, submission rates, and data quality. Useful for programme management and internal accountability.
      </Card>

      <Card title="Donor Report" icon="handshake">
        A funder-facing narrative that maps field data to grant indicators and logical frameworks. Formatted for external distribution.
      </Card>

      <Card title="Portfolio Report" icon="chart-bar">
        An aggregated view across multiple projects. Available on Growth and Enterprise plans.
      </Card>
    </CardGroup>
  </Step>

  <Step title="Data quality check">
    Before generating, Terriqon runs an automated data quality check on the underlying submissions. If the data is insufficient — too few submissions, too many blank required fields, or a data quality score below the threshold — you'll see a **quality warning**. You can proceed to generate anyway (the report will note the limitation) or return to your data and follow up with your field team first.
  </Step>

  <Step title="AI generates the draft">
    If quality checks pass, the AI begins generating the report. This typically takes **under 15 minutes**. You can navigate away and continue other work — Terriqon runs the generation in the background.
  </Step>

  <Step title="Receive the notification">
    When the draft is ready, you receive both an **in-app notification** and an **email**. Click the notification to go directly to the review screen.
  </Step>
</Steps>

## Reviewing the Draft

The report review screen is split into three panels:

* **Report narrative** (left): The full AI-generated text, broken into clearly labelled sections (Executive Summary, Activities, Outcomes, Recommendations, etc.). Every section is editable.
* **Data metrics** (right): The underlying aggregated figures used to generate the narrative — submission counts, completion rates, field-level statistics, and any computed KPIs.
* **Redaction log** (bottom): A record of every field that was automatically stripped from the AI pipeline because it was tagged as Personally Identifiable or Restricted. Review this log to confirm PII exclusion before approving.

Work through each section of the narrative, cross-referencing the data metrics panel to verify accuracy.

## Editing the Report

Every section of the report narrative is editable inline. Click any paragraph or heading to activate the text editor for that section.

Common edits include:

* Adding programme context the AI couldn't infer from submission data alone
* Adjusting tone to match your organisation's voice or a specific funder's expectations
* Correcting figures if a data anomaly is spotted
* Inserting a quote or case study from your field team

All edits are tracked in the **Audit Log**, which records what was changed, by whom, and at what time. The audit log cannot be edited or deleted.

<Note>
  Once approved, the report is final. If you need to make changes after approval, create a new report version from the Reports tab rather than attempting to modify the approved document.
</Note>

## Approving a Report

When you are satisfied with the draft:

<Steps>
  <Step title="Click Approve Report">
    Click the **Approve Report** button in the top-right corner of the review screen.
  </Step>

  <Step title="Confirm with your name">
    A confirmation dialog asks you to type your name as a digital signature. This confirms that a named, authorised individual has reviewed and approved the content.
  </Step>

  <Step title="Download or share the report">
    The approved report is now available for download. Export formats include **CSV**, **Excel**, and **JSON**. The report carries a **verified status badge** showing the approver's name, role, and timestamp.
  </Step>
</Steps>

The approval is permanently logged and cannot be removed. This creates an auditable chain of custody from raw field submission to approved funder deliverable.

<Warning>
  Do not share an unapproved AI draft externally. Approved reports carry a verified status badge and can be independently audited. Unapproved drafts carry no such verification and could expose your organisation to compliance risk if shared with funders or partners.
</Warning>

## Rejecting a Report

If the draft is inaccurate, incomplete, or otherwise unsuitable for approval:

<Steps>
  <Step title="Click Reject">
    Click the **Reject** button in the top-right corner of the review screen.
  </Step>

  <Step title="Enter a written reason">
    A reason is **required** — type a clear explanation of why the draft was rejected (for example: "Submission count is too low for the reporting period; please collect additional data from the northern district before regenerating"). This reason is visible to all Admins and Managers on the project.
  </Step>

  <Step title="Confirm rejection">
    Click **Confirm Rejection**. The report returns to **Draft** status. The rejection reason and timestamp are logged permanently in the Audit Log.
  </Step>
</Steps>

Once rejected, you can either regenerate the report after addressing the underlying data issues, or edit the existing draft manually and re-submit it for approval.