Mindcraft: Welcome to the era of Cognitive Behavioral Scoring

AI is perhaps the most over-hyped terminology used in financial technology circles. So much so, that a lot of industry veterans snigger and snort whenever artificial intelligence is mentioned. There’s good reason for that skepticism. The industry is teeming with vendors that claim to employ artificial intelligence, when all they’re doing is reaching decisions using expert rules and manual data analytics.

Now data analytics and expert rules are immensely helpful, and when done properly they yield high accuracy in terms of credit scoring. But the trouble with these solutions is, to yield results you are either heavily depended on your solutions provider, or you must hire your in-house data scientist. This results in high cost, and longer processing times.

This is where Mindcraft succeeds, while others fail.

We have been perfecting the technology for over a decade, and it has become one of our core offerings. Two of our clients, Malaysia’s two prominent banks, have been using Mindcraft for behavioral scoring and early (up to 6 months ahead) warning for NPLs and for identifying “self-curing” customers. Self-curing customers are delinquent customers who will end up fixing their NPL situation without any prompts from the bank. Identifying self-curing customers can save the bank valuable time and resources.

Why Mindcraft is in a Class of its Own

Mindcraft has out-of-the-box capabilities such as machine-learning, cognitive behavioral scoring, and continuous and autonomous self-learning to improve accuracy.

Imagine you are using a market-standard behavioral scoring and data analytics solutions provider. Each time you introduce a new data point, for example a customer’s social networking profile data and try to integrate it with other available datasets such as CRM data or transaction behaviour, you can’t do it yourself unless you have specific set of data science related skills. So, you end up calling your solutions provider, asking them to integrate this new data point then playing the waiting game.

You have to wait until they are done implementing, which of course will cost extra, and you can’t see results until they take their sweet time incorporating the new dataset. Your solutions provider has you in a hostage situation, you’re at their mercy.

Now if you have Mindcraft, you can input the dataset yourself. You can compare the new challenger model with your existing champion model and set it to work for you immediately. And over time, if you select this option, Mindcraft will autonomously replace the existing predictive model with the one that yields better accuracy, with human agreement if necessary.

As for behavioral scoring, Mindcraft uses a cognitive model;an approximation to human cognitive processes that are used for comprehension and prediction. It can use your existing dataset, add new data points and rules anytime, autonomously approve or reject loan applications based on the confidence score. Let’s say if Mindcraft is 90% confident that Mr. Joe Good Customer with A1 credit rating will pay back the loan without complications, the application can be approved instantaneously without human involvement.

In some cases, a person who has applied for a loan may not have sufficient credit history for autonomous approval. In those cases, Mindcraft will still generate a social credit score, using available social networking data like that person’s LinkedIn, Facebook and Instagram profiles, and if available; smart-phone usage data. Once the social credit score is generated, it will be paired with a confidence score, based on which a human can approve or reject the loan.

The best part about Mindcraft is that its decisioning is not black-boxed. Mindcraft is an explainable AI, which means that it can provide an explanation behind each decision reached.

While Mindcraft is a great behavioral scoring and early-warning system, the AI engine can be used to calculate a potential customer’s propensity to apply for a car loan, based on your existing database of every customer who has ever applied for car loans. Mindcraft will identify common patterns of behaviour so you can cross-sell/up-sell to existing customers.

Mindcraft can also help expand your existing customer base by identifying potential/future customers who are most likely to apply for a car loan.

With Mindcraft you can reach the right customer, at the right time, with the right product.

Experience a higher ROI, and lower marketing costs.

In a nutshell, Mindcraft can autonomously handle:

1. Consumer onboarding

2. Behavioral scoring

3. Loan approval/rejection

4. Self-tweaking champion/challenger models for higher accuracy

5. Conduct analytics and include new data points without the help of a data scientist

At JurisTech, when we use the term “Artificial Intelligence”, we really mean it.

To learn more about Mindcraft, click here.

By | 2018-05-17T16:34:22+00:00 7th May, 2018|Artificial Intelligence, Featured, Insights|