Reporting intelligence
AI Custom Reports via Prompts
Users can ask natural-language reporting questions, generate permission-aware outputs, get AI summaries/highlights, save prompts, re-run saved prompts, and schedule saved prompts where feasible.
Natural languageAI report summariesSaved promptsScheduled prompts
1. Feature Overview
Prompt-based reporting reduces the need to hardcode every possible report. It converts user questions into supported filters and queries over available Journey CRM data.
Prompt: Show hot leads with no follow-up in the last 48 hours.
AI Summary: 12 hot leads need follow-up. 4 have re-tour activity and 3 are from high-value referral sources.
Next Step Suggestion: Prioritize the 4 re-tour prospects and assign overdue follow-up tasks today.
Full Results: Open full list view for all matching records.
2. Workflow
flowchart TD
A[User enters prompt] --> B[AI interprets intent]
B --> C[Validate against reportable CRM data]
C --> D[Apply permissions + property filters]
D --> E[Generate query / filters]
E --> F[Return result count + table]
F --> G[AI summary + highlights]
G --> H[Next-step suggestions]
H --> I{User action}
I -->|Open| J[Full list view URL]
I -->|Save| K[Saved Prompt]
I -->|Schedule| L[Scheduled Prompt]
I -->|Export| M[CSV / Excel / PDF]
3. Planned Capabilities
- Prompt-based report request.
- AI converts question into supported filters/query.
- AI summarizes the report output and key trends.
- AI provides highlights and next-step suggestions.
- Long reports show a summary inline and provide a URL/link to the full list view instead of rendering every row inline.
- Save frequently used prompts.
- Re-run saved prompts.
- Schedule saved prompts/reports where feasible.
- Export output where supported.
- Respect user permissions and property/PMC access.
4. Example Prompts
| Prompt | Expected Output | AI Highlight |
|---|
| Show hot leads with no follow-up in the last 48 hours. | Prospect list view. | Top overdue hot leads and suggested follow-up priority. |
| Show re-tour prospects by property this month. | Property-level re-tour table. | Properties with strongest second-visit intent. |
| Show post-tour surveys with negative sentiment. | Survey response report. | Common objections and coaching themes. |
| Show referral sources by conversion rate. | Referrer/source performance table. | Best and weakest referral sources. |
| Show leads who asked about pricing in the last 30 days. | Intent-based prospect report. | Pricing interest by property/source. |
5. User Cases
Long result set handled as list view
Scenario: User asks for all prospects with no activity in 30 days and gets 400 results.
System behavior: AI provides summary/highlights inline and links to the full filtered prospect list.
Result: User gets insight immediately without overwhelming the chat/report output.
Saved report prompt
Scenario: Manager frequently checks hot leads with no recent follow-up.
System behavior: User saves the prompt and re-runs it weekly.
Result: Repeat operational reporting becomes self-service.
Scheduled executive prompt
Scenario: Executive wants a weekly re-tour snapshot.
System behavior: Saved prompt runs on schedule where enabled.
Result: Leadership receives consistent insight without manual report setup.
6. Data Boundary and Governance
- Reports can only use available Journey CRM data.
- Prompt execution must respect permissions.
- AI should not expose cross-PMC/community records.
- Saved prompts store owner, scope, filters, and schedule settings.
- If a prompt cannot be answered reliably, system should explain the limitation.
7. Acceptance Criteria
- User can ask a natural language reporting question.
- System returns a report summary, highlights, and table/list output where supported.
- Long reports provide a link to full list view.
- User can save and re-run prompts.
- Scheduled prompt support is available where prioritized.
- Prompt results follow user permission and property access rules.