Airbyte + Neon + Python + Vercel
A simple cloud dashboard that shows which customers may leave, why they may leave, and how much monthly revenue is at risk.
Click any number card to see what it means.
| Customer | Chance of Leaving | Risk Level | Monthly Bill | Contract Type | Months as Customer |
|---|---|---|---|---|---|
| 5178-LMXOP | 86% | High | $95 | Month-to-month | 1 |
| 7216-EWTRS | 86% | High | $101 | Month-to-month | 1 |
| 5567-WSELE | 85% | High | $95 | Month-to-month | 3 |
| 7024-OHCCK | 85% | High | $94 | Month-to-month | 2 |
| 9300-AGZNL | 84% | High | $94 | Month-to-month | 1 |
| 8098-LLAZX | 84% | High | $95 | Month-to-month | 4 |
| 1374-DMZUI | 84% | High | $94 | Month-to-month | 4 |
| 9497-QCMMS | 84% | High | $94 | Month-to-month | 1 |
| 4424-TKOPW | 84% | High | $94 | Month-to-month | 2 |
| 2720-WGKHP | 84% | High | $94 | Month-to-month | 2 |
Longer bars mean stronger reasons. Click a bar to see what it means.
This section explains the table and card names in plain language.
CustomersHow many customer records are in the clean data.
Customers who really leftThe percent of customers who actually stopped service in the data.
Average chance of leavingThe model's average predicted chance that customers may leave.
Likely to leaveCustomers whose predicted chance of leaving is 70% or higher.
Monthly bill at riskMonthly money connected to customers who are likely to leave.
Model scoresSimple checks that show how well the prediction model worked.
CustomerThe customer's unique ID.
Chance of LeavingThe model's estimated chance that this customer may leave.
Risk LevelA simple label: Low, Medium, or High.
Monthly BillHow much the customer pays each month.
Contract TypeThe customer's plan, such as monthly or yearly.
Months as CustomerHow long the customer has stayed with the company.
ReasonA customer detail used by the model, such as contract type or service.
StrengthLonger bars mean the reason matters more to the model.
DirectionRed raises the chance of leaving. Green lowers it.