MAIBE


Machine Learning trained on past financial transactions,
to predict the next most probable transactions

Smart Banks

Despite the financial sector being the second largest in the S&P 500 Sector Weightings, at #49 there is only one financial services company listed in the MIT Technology Review latest list of the 50 Smartest Companies

Better financial decisions

MAIBE predicts for each user the upcoming most probable transactions. In line with behavioral economics theory by the winner of the The Nobel Prize in Economics Science 2018, this visibility into future expected transaction provides a choice architecture that could nudge users into better financial decisions

In a production environment, a machine learning model typically learns from millions of example data items and produces inferences in hundreds to less than 20 milliseconds

Machine Learning with Amazon SageMaker

MAIBE is trained on munged financial transaction data, no personal information whatsoever, under highly sophisticated data models tightly coupled to its machine learning algorithms.

The result is magic! It wows customers my predicting their next transaction, saving them time, and providing them with an outlook of the next expected upcoming transactions

Predicting the next most probable transaction

Deliver a jaw-dropping user experience by anticipating the customer's next transaction.

Available as a RESTful API:
/maibe/accounts/{anonymousAccountID}/channels/{channelID}/probabilities

ATM


Do you wish to withdraw 150€?

Mobile App


Do you want to make the 500€ rent payment?

On-line banking


Do you want to process payroll?

Chatbot


Payment from customer has posted

Can receive two possible answers from customer

Yes!

Wow, how did you know? Engage!

No

Please show me the regular menu

Predicting low balance situations

In addition to anticipating one transaction, MAIBE can anticipate all the transactions in the near future and warn of low balance situations