Fair Lending and Loan Predictions

Determine lending risk and payment default rates in situations with limited or non-existent credit bureau utilization

Simplify the process of predicting whether the customer will default on their credit or not, minimizing losses. 

Accurate models that will predict risk profiles for customers based on their Customer 360

Model Explainability that provides reasons behind the predictions of deployed models

Generate profits for the institution through differential borrowing/ lending rates.

H2O.ai saved $20M per year in credit underwriting for single customer

More accurate predictions of bad credit and with high probability of defaulting

Increased accuracy in interest rates and expected returns for each customer, according to their risk profile

Create models and use custom H2O recipes specifically built for generating features/variables that provide associated information about defaulting behavior. 

Make data available to the credit officer who can further slice and dice the data and consume the information intuitively.

H2O AI App built with Wave, an SDK for Data Scientists, can perform:

• Data Preparation

• Non-algorithmic modeling

• Supervised Machine Learning

• Interactive Component: users can upload the training dataset to train the model first, then upload the testing data and view the results.

• The results viewed could be in tabular form, grouped results ( all defaulters, people of xyz location,etc) or a dashboard for a single customer.

• Dashboard to display insights like percentage chance of defaulting and graphs ( eg - scatter plot of all customers, position of selected user wrt others), the rate of interest charged/the return expected for specific customer, and a textual insight explaining the prediction for better transparency.