|
|
|
Scorto™ Model Maestro
|
Scorto Model Maestro is a predictive analytics software for the development of scoring models and loan portfolio analysis. It allows identifying the key factors, impacting customer’s creditworthiness, developing scorecards, analyzing scorecards efficiency and exporting scorecards to the decision management system.
|
Application
Behavioral
Collection
|
Fraud detection
Custom-purpose
scorecards supported
|
|
- Extremely easy to use for a business user.
-
Since Model Maestro is wizard-based, there is no need to learn command-line language. Everything is configurable via wizards and at-hand properties. For example, you do not need to write a script or create inputs for Trade-Offs Chart creation as well as for any other type of graph related to scorecard performance evaluation. All relevant graphs, curves, tables, reports are pre-defined for you and just one-click away. You can view the related graphs simply clicking on certain parameters, like axes or normalization or log scale view, etc.
And since the tool is created purely for scorecard development, there are no extra belts and whistles to learn, no obsolete functionality, no universal approaches a la “one fits all tasks”, everything is done just for scorecard development.
Scorto™ Model Maestro has an extremely small learning curve, saving your time for scorecard development and making your work with the tool really enjoyable.
- Complete solution for scorecard development.
- Up-to-date with artificial intelligence ready to back you up if required.
-
Not only logistic regression supported, but Neural Networks, Decision Trees, Decision Rules and claster analisis for the cases when the model’s performance can be improved with non-linear modeling techniques and advanced artificial intelligence. Every technique and its parameters are explained in wizards, all parameters are initially predefined using proprietary best parameters identification algorithms, so you don’t need to be a quant to create your neural network based scorecard.
- Best practices incorporated.
-
- - How to deal with missing values, outliers, wrong type values in your portfolio?
- - What to do if you have small amount of historical data?
- - What is the best modeling technique for your case?
- - How to demonstrate the power of your scorecard to all stake-holders?
- - Will scorecard be as good with new data as with training dataset?
- - Can Gini be improved?
Answers to all these and other questions are incorporated in Model Maestro. So, the best practices previously available only to the top-notch experienced scorecard developers inside credit bureaus are now at your service, predefined for you.
- Basel II ready.
From business perspective your risk managers and scorecard developers will have:
- The power to create scorecards in-house, keeping all the knowledge and control with you, inside your organization, without the need to transfer sensitive customer data to 3rd party scorecard developers
- Ability not only to create, but monitor and improve your scorecards with pre-defined reports
- Fast adaptation to changing market conditions
- Fast deployment of new strategies, rules and predictive models
Model Maestro demystifies high science of scorecard development algorithms, brings it down to your practical needs, with standardized pre-defined development flow and wizard-based interface.
|
|
|
|
|
|
This is not just a tool, but a solution.
Purchasing Model Maestro you get 2-day on-site training with your data for free. So in the end you have a perfect solution for your problem: trained employees, configured software and support.
|
|
Any custom report to your specifications can be developed within 2 weeks.
Custom version within 1-3 months with specific or simplified functionality. Explain us your scorecard development process and we will tailor Model Maestro to fit it.
|
|
Really fast support.
Scorecard development specialist available next day after your request to help you create, improve or audit your scorecard or just help you with certain data preparation or model valuation step.
|
|
|
|