Skip to main content
AI Assistant

​
Natural Language Analytics

Instead of building manual reports with filters and pivot tables, you can ask Parsa AI direct questions about your CRM data in plain English. The AI translates your query into a data analysis, returns results as a chart or table, and you can save it for later.Go to AI → Analytics or use the AI chat panel with any analytics-style question.
Example Queries

​
Example Queries

​
Pipeline and Revenue

What's the total pipeline value by stage?

Show me monthly closed won revenue for the last 6 months.

What's the average deal size for each rep?

Which opportunities are most at risk of slipping this quarter?

​
Leads and Contacts

Show me leads created in the last 30 days, grouped by source.

What percentage of leads converted to contacts last month?

How many new contacts were added each week this quarter?

Which lead sources produce the highest conversion rate?

​
Activities and Engagement

How many emails were sent by each team member last week?

Which contacts haven't had any activity in the last 45 days?

What's the average time between lead creation and first contact?

​
Cases and Support

Show me open cases by priority.

What's the average resolution time for cases this month?

Which accounts have the most open cases?
Dashboard

​
Saving Charts to Dashboard

After running an analytics query, you’ll see the result as a chart or table. To save it:
  1. Click Save to Dashboard below the chart.
  2. Give the chart a name (e.g., “Pipeline by Stage — Live”).
  3. The chart is added to your CRM dashboard and refreshes automatically with live data every time you view it.
You can add multiple charts to your dashboard and rearrange them by dragging. Remove a chart from the dashboard by clicking the × icon in dashboard editing mode.
Reports

​
Saving to Reports

For analytics you want to revisit regularly without pinning to the dashboard:
  1. After running a query, click Save to Reports.
  2. Give the report a name and optional description.
  3. Access saved reports from AI → Reports at any time.
Reports re-run against live data each time you open them. You can also export any report to CSV by clicking Export from the report view.
Advanced Queries

​
Combining Filters with Natural Language

Show me closed won deals in Q1 where the assigned rep is Alex or Jordan.

Which leads created via the Website Form source in the last 60 days are still in New status?

What's the pipeline value for accounts in the Technology industry with more than 50 employees?
The more context you provide, the more precise the results.
AI analytics run against your live CRM data. Deleted records are not included in query results. If a chart looks off, verify that the underlying data (e.g., close dates, stage values) is accurate in the CRM.
Tips

​
Tips for Better Queries

  • Be specific about time ranges — “last 30 days” or “Q1 2025” produces better results than “recently.”
  • Specify groupings — “by rep,” “by stage,” “by source” tells the AI how to slice the data.
  • Ask for comparisons — “compared to last month” or “vs. the same period last year” triggers trend analysis.
  • Use CRM terminology — Terms like “closed won,” “lead source,” and “pipeline stage” are understood natively.

​
Pipeline Health Queries

Ask about pipeline movement, not just snapshots. Behind the scenes the analytics engine reads the OpportunityStageHistory ledger — one row per stage transition — so it can answer about drift over time.
“What’s our average cycle time from Prospecting to Closed Won in the last 90 days?” “Which stage do we drop the most opportunities from?” “How many opportunities sat in Negotiation longer than 14 days last quarter?”
The chart shows drop-off bars per stage, an average-cycle-time line, or a velocity-by-stage breakdown depending on what you asked. Save it as a Pipeline Health widget on a report to track it over time.

​
Automation Queue Queries

The analytics engine knows about automation-spawned tasks too (via the automation_action field on Task). So you can answer:
“Which automation rule has fired the most this month?” “How many automation send_email actions failed in the last week?” “Show me the time-to-complete distribution for create_task actions.”
Useful for tuning automations — if one rule is doing 80% of the firing, it might be configured too aggressively; if half its actions are failing, something’s broken with the rule’s email settings.