Product | CREDI AI

Make More Good Decisions and Fewer Bad Loans

Enhance segmentation, pricing and collections

Aligning AI strategy with overall lending strategy and goals.

Identifying the goals and selecting KPIs will help you evaluating performance for AI investment.

Some of the most common goals and results our customers are using AI for include :

High-performing predictive models anytime you need. Model development from start to production in three (3) months.

1. Data Development: To build a model, CREDI AI requires 1,000 records of previously issued loans with a status, whether good or bad. You are not required to pre-process data.

2. Model Building: Based on the analysis of your business and data, we build a customise model that best suits for your needs by using the most appropriate algorithms, such as decision

3. Validation: CREDI AI provides a validation report after model building so that what features are being selected as key variables and the accuracy of the model

4. Deployment: CREDI AI builds a API module that can be easily integrated into your operation system and workflows. It’s ready for real-time predictions.

5. Model Monitoring: The module monitoring and modelling services will allow you to quickly react and update models when required. Easily adapt to any changes within your

Does CREDI AI utilize alternative data?: No. CREDI AI works with whatever the data you provide. We are specialized in pattern detections. It does not use data from other sources when it comes to build AI credit scoring models based on your loan performance data. However, we help customers setting up a structure to start collecting alternative data. For more information.

Are your models black boxes?: No. Our model is understandable and explainable. It’s true that the models built by using non-parametric classifiers are harder to explain. However, we still provide the stats required to validate them. See more details.

What is your price structure?: We normally offer an initial set-up fee and monthly subscription structure based on your volume of loan applications. A minimum monthly subscription fee might be applied if application volume is too small. You can also check more details about system integration how machine learning models will be integrated into your existing operating system.

Want to launch a new lending operation, but have no historical data. Can you still help?: Yes. We’ll set up a structure that will allow you to leverage machine learning as soon as you collect enough data to create your own models.

What data will be required for CREDI AI to build a risk prediction model?: CREDI AI normally requires at least 1,000 loan performance data whatever you have. However, we don’t require any kinds of personal information which specifies individuals. Plus, data anonymization is applied when required. For more details.

How does CREDI AI protect and secure customer data?: We have set data security methods, practices and processes as strong as other data companies do such as encryption, multi-factor authentication, access control, network security, activity monitoring etc. Normally we encrypt all of the customer data which will be shared between lenders and CREDI AI. Application data will be anonymised when needed. Also we build a high security environment to use dedicated server networks which are assured to be protected from the information leakage or misuse.