We’re excited to announce the launch of our new Payroll Prediction Model dashboard!
This dashboard allows you to monitor the accuracy of our Payroll Prediction Model, which detects payroll in bank transactions and predicts the next paycheck date and amount with over 90% accuracy.
Detecting the next payroll date for a user is challenging because payroll transactions often have generic descriptions, lack payroll-specific keywords, and have varying cadences.
Our Payroll Prediction model has allowed our customers to reduce NSFs by up to 30% and improve customer experience by aligning the collections with an accurate pay date.
Payroll Prediction Model Dashboard
This dashboard reveals insights including:
- Number and percentage of your users with income (See the full list of income types on our docs)
- Number and percentage of your users with payroll
- Accuracy of our payroll prediction model
- User examples to demonstrate our data quality
Check out this video which gives an overview of the dashboard:
The Payroll Prediction Model enables you to:
- Understand a user’s payroll frequency
- Predict the payroll’s next_date and next_amount
Use Cases
- Cashflow underwriting
- Optimizing loan repayment schedules
- Predicting overdrafts
- Reducing NSFs
- Building cashflow forecasting tools
How the Payroll Prediction Model Works
- Pave’s matching algorithm accurately labels and clusters payroll transactions
- The model predicts the next paycheck date and amount with over 90% accuracy
- Pave’s human-in-the-loop model ensures we’re able to clean and tag payroll transactions with >90% accuracy
For more information, check out our docs.
Reach out to howdy@pave.dev if you’d like to explore this dashboard, and we’ll provide access!
