Analytics Engines

Mitek’s software is delivered as analytics engines. These engines are based on innovative analytics technologies and unique algorithms. The engines can be integrated with most IIoT software environments. In a proof of concept demonstration, the engine can run as a stand-alone application to quickly get the added value results.

How Does Analytics Engine Work?

Analytics engine is a stand-alone program that reads files from the input data pool, processes them, and generates output data files. The analytics engine might include predictive and/or prescriptive analytics. Its operation can be tuned through JSON configuration data files. The engine uses MapReduce for scaling to big data sets and large computing clusters. In a stand-alone deployment for a proof concept demonstration, Mitek’s report generator converts engine output data into value-added portable reports.

Our analytics engines are developed following the best systems engineering practices for critical applications. The process includes rigorous verification and validation. In some applications the existing engine just needs to configured for deployment; in others, the prescriptive analytics is customized. The predictive analytics can integrate existing knowledge from engineering and data science work, e.g., in the form of models. The analytics engines have open architecture and can be integrated with existing analytics systems.

Data Science Integration

Most Big Data work is related to web applications that might support millions of users. Data Scientists are focused on knowledge discovery. They use Machine Learning tools for building descriptive models for on-line analytics. This research and exploration focus means there is a steep learning curve for each new analytics deployment. Mitek’s focus is on deploying analytics in mission-critical applications of the Industrial IoT. Deploying our analytics engines helps to span the gap between the knowledge and operationally deployed applications; eliminate mistakes in mission-critical environment. The high value of the applications and high cost of the mistakes justifies the use of advanced analytics engines in the Industrial IoT. The analytics engines can integrate and scale up can scale up the existing baseline models to the IIoT Big Data.