Test environment. No real payments are taken. “Demo mode” runs use rule-based text extraction, not AI.
Atlas Automation GroupSign inGet started

All agents · Customer support and success

Customer Health and Churn Review

Beta

At-risk accounts flagged from your usage and ticket export, with risk levels computed by fixed rules in code and a plain-language reason and next step for each.

Try Customer Health and Churn ReviewPricing

How it works

Paste or upload a CSV with one row per account (name plus any of: usage this and last period, support tickets, last active date or days since, NPS or CSAT). Risk levels come from transparent rules in code; the AI only explains them and suggests next steps.

What you give it

  • Pasted CSV
  • .csv

Limitations

  • Risk levels are rule-based (usage drop ≥30%, many tickets, inactive >30 days, low NPS/CSAT); they are signals for follow-up, not predictions.
  • Column roles are detected from header names; check the 'columns used' table.
  • Up to 10,000 accounts and 60 columns.