Ai system for processing historical data to identify and reconfigure prescheduled, suboptimal events

A method for leveraging an artificial intelligence (“AI”) engine to intercept and refine a generic notification generated by a platform for transmission to a plurality of customers, the AI engine using machine learning and historical behavioral trends, the method comprising:

• using the platform, scheduling the generic notification for deployment to the plurality of customers at a first time of day and in a first method of presentation, the generic notification including first information, the first information including data relating to a savings opportunity and a functionality linked to an account of a recipient of the notification to pay off outstanding debt, wherein each of the plurality of customers are determined by the platform to have an account that satisfies one or more predetermined conditions;
• periodically monitoring, using the AI engine, randomly-selected historical customer data stored by the platform and associated with a customer included in the plurality of customers, the randomly-selected historical data having been generated within a predetermined time window and including notifications transmitted by the platform to the customer, wherein a size of the randomly-selected historical customer is less than a predetermined threshold value and the notifications reviewed by the AI engine during the monitoring are a subset of notifications stored by the platform;
• determining, using the AI engine, during the monitoring and prior to the transmission of the generic notification to the customer, that the notifications monitored by the AI engine failed to elicit a positive response from the customer, the determination prompting the AI engine to access and review a larger amount of customer data than was accessed during the monitoring, the AI engine, when prompted, also being configured to:
• instruct the platform to put the scheduled deployment of the generic notification to the customer on hold;
• tag notifications stored in the database that elicited a positive customer response;
• filter the tagged notifications by a time of day, a method of presentation, and a topic;
• generate a predictive recommendation for a targeted customer notification, the targeted customer notification including an optimized time of day, an optimized method of presentation, and the first information;
• determine, using the AI engine, if the first time of day is the optimized time of day and if the first method of presentation is the optimized method of presentation;
• when the AI engine determines that the first time of day is the optimized time of day and the first method of presentation is the optimized method of presentation, instruct the platform to remove the hold on the generic notification and to schedule the generic notification for deployment to the customer;
• and
• when the AI engine determines that either the first time of day is different from the optimized time of day or that the first method of presentation is different from the optimized method of presentation, the AI engine being further configured to:
• overwrite for the customer, and not for the plurality of customers, the generic notification with the targeted customer notification;
• and
• schedule for deployment the targeted customer notification at the optimized time of day in the optimized method of presentation, the targeted customer notification including the first information;
• wherein:
• the periodic monitoring of the AI engine of the historical customer data enables the platform to integrate intermittent AI decisioning into customer interactions;
• and
• the AI engine is configured to monitor, at a given point in time, historical data associated with less than a predetermined number of customers to ensure reduced, relative to a predetermined baseline, processing usage.

A system using light-weight computing to generate and refine generic notifications using artificial intelligence (“AI”), the system comprising:

• a platform that, in operation:
• generates a generic notification for transmittal to a plurality of customers at a first time of day and using a first method of presentation, the generic notification including first information;
• and
• provides an interface for interacting with an AI engine;
• a database for storing historical customer data, the historical customer data including notifications transmitted by the platform to the plurality of customers and responses of the plurality of customers to the notifications;
• an application server running the AI engine for:
• periodically monitoring, at a given point in time, historical customer data relating to less than a predetermined number of customers included in the plurality of customers to ensure reduced, relative to a predetermined baseline, processing usage;
• determining, during the monitoring, that notifications transmitted by the platform to a customer included in the plurality of customers have failed to elicit a positive response from the customer, the determination prompting the AI engine to access and review a larger amount of customer data associated with the customer than was accessed during the monitoring;
• and
• in response to the determining, instructing the platform to put the transmission of the generic notification to the customer on hold;
• the platform for receiving the instructions from the AI engine and putting the transmission of the generic notification to the customer on hold;
• the application server running the AI engine for performing machine-learning analysis on data stored in the database and associated with the customer to generate a targeted customer notification, the machine-learning analysis including:
• tagging notifications stored in the database that elicited a positive customer response from the customer;
• filtering the tagged notifications by a time of day, a method of presentation, and a topic;
• generating a predictive recommendation for the targeted customer notification, the targeted customer notification including an optimized time of day, an optimized method of presentation, and the first information;
• determining if the first time of day is the optimized time of day and if the first method of presentation is the optimized method of presentation;
• when the first time of day is the optimized time of day and the first method of presentation is the optimized method of presentation, instructing the platform to remove the hold on the generic notification and to schedule the generic notification for deployment to the customer;
• and
• when either the first time of day is different from the optimized time of day or the first method of presentation is different from the optimized method of presentation, instructing the platform to:
• overwrite the generic notification with the targeted customer notification for the customer, and not for the plurality of customers;
• and
• schedule for deployment to the customer the targeted customer notification at the optimized time of day in the optimized method of presentation, the targeted customer notification including the first information;
• wherein:
• the first information includes data relating to a savings opportunity for presenting to the customer and a functionality linked to an account of the customer to pay off outstanding debt, the generic notification being generated by the platform for the customer when the platform determines that an account of the customer satisfies one or more predetermined conditions.