ML21™ Machine Learning Platform

CodeNext21′ ML21™ platform for is suitable for finding anomalies in data and can be applied to different types of data.

Anomaly detection explained in “Jip & Janneke language”
An anomaly, also known as an “anomaly,” is a data point that is out of line with previously observed data. If a graph of the data shows a nice line, then you can consider data points that are not on that line as anomalies. However, most data is not too easy to contain in one line (also called a “dimension”). The data companies process sometimes contains as many as millions of data rows, and each row can contain tens to hundreds of dimensions. Such quantities can no longer be analyzed with the naked eye.

Machine Learning makes it possible to automatically find anomalies and do analysis on the data. However, there are many ways to apply ML and look at the data. Depending on what you want to achieve, different techniques and statistical algorithms are available. CodeNext21 has developed new algorithms for its platform and combined them with a unique scalable ML platform design..

Our proprietary machine learning algorithms are “clearbox” and real-time.

ML21™ can be applied for various purposes such as anti-fraud solutions (such as credit card fraud, payment fraud or money laundering fraud) or finding anomalies in processes of manufacturing or logistics. Our platform provides a clear explanation of why the system alarms which enables you to communicate clearly and transparently with your customer.

  • High performance for real-time processing of large amounts of data
    ML21’s algorithms are specifically designed for high performance. In addition, ML21 scales very well, allowing real-time analysis of several thousand records per second.

  • OEM to implement (for example, for ERP software)
    You can use our ML engine through an API service, in the cloud or self-hosted. We provide our services to software developers who want to implement ML services to increase value from data.

  • Clearbox: So clear explanation of outcomes
    The outcomes of our platform are given in a traffic light principle: Red (Deviation), Orange (Suspected/Possible deviation), Green (Normal);
    Each outcome gets a clear motivation about the conclusions the platform makes. We do this with graphic support. You don’t have to be a data analyst to use our platform!

  • Implementation in less than 8 weeks
    Almost all ML models require a training period to learn a model. If the assumptions of the model change, the model must be re-trained. So the solution is a static ML model.
    Our platform immediately starts “learning” when feeding data. It in not necessary to train the model first with a static model. This offers some major advantages over many other ML solutions. Thus, our model does not need to be re-trained when models change. Even if data changes over time, the model can learn from the new data and your feedback.

  • High Quality
    Quality is a flexible term. When analyzing your data, you want to find anomalies that you find interesting because they are actually an anomaly. Moreover, you do not wish to overlook important issues. Our technology applies multiple algorithms to your data to deliver the highest possible quality. When we talk about quality we say “Our ML finds anomalies that matter”!

Choose the right path for the future

Brainstorming about applying our technology to your data? Curious about how your organization can handle data better? Contact us.