Predictive AI Applications
Build Predictive AI Applications that leverage your big data with a variety of algorithms. Create large-scale omni-channel applications such as Customer Journey Analytics across multiple customer interaction platforms.
Use Foxtrot Code to create Predictive AI Apps that pipe big data sources directly into Machine Learning Algorithms. Then use the predictive results to create Prescriptive Solutions that intervene directly with your customers, machinery, and other processes.
Our featured application predicts kWh usage based on temperature. The demonstration leverages over 65 million Smart Meter readings over a three year period. We provide a complete set of analytic applications including data integration, statistics, clustering, and model fitting. The application demonstrates the results over 7 different machine learning algorithms.
The Smart Meter application is a typical machine learning application. There are phases of reduction and enrichment, followed by calculation of statistics, clustering, model fitting, and model application. The phases feed each other, and they provide the capability to run each phase in production as separate 24×7 batch tasks to meet the requirements for volume and velocity of the application in production.
You can download and reuse the application for you own needs. We have also provided a dashboard to review the predictive value of the algorithms we featured in the demonstration.
You can access the live dashboard by clicking HERE the login for the site is username: guest password: guest.
Built-in Machine Learning
With Foxtrot Code you can Fit machine learning models, save the models, and then Apply the models in real-time, and at large-scale to incoming data. Use Splunk’s built in machine learning algorithms (below), or bring your own. Foxtrot Code can also pipe refined data to external machine learning applications running on graph databases such as: NEO4J, Watson, Objectivity, SparkML.