AnoFox
The fox has an eye on your business. It provides flexible components for anomaly detection and timeseries prediction. It is easily managable as part of your cloud data plattform.
Companies collect more and more data stemming from an ever growing variety of sources. But that does not automatically mean, that business is able to make use of data or derive predictions from it. AnoFox provides components companies can incorporate in their cloud base data plattforms, to accelerate their journey to business value from data.
AnoFox helps to monitor and predict increasing number of data streams or time series from IoT and business data. It provides automated, unattended, online prediction and anomaly detection for a large number of time series in parallel. We use a combination of robust established statistical methods combined with state of the art machine learning methods to analyze the incoming data streams in realtime. A supervision algorithm combines the predictions of the different methods and provides top notch anomaly scores and future predictions.
As companies get more and more data driven, the platforms they use to gain insight or drive their ditigtal business get more and more important. We therefore think that owning and having control over this platforms is critical for our customers future success. Therefore we don't provide our solutions as a hosted service, which would require to ship the data to our systems. Rather we provide components the customer can deploy as solution accelerators in their virtual private clouds.
Often times it is hard to find real business value when bringing machine learning to the enterprise.
The breakthroughs of big internet companies in deep learning, which received a lot of press coverage,
apply to a very narrow set of domains. At the end, there is limited business value for a typical business
in beeing able to detect and classify object on photos from mobile phone cameras.
In contrary there is a mutlitude of use-cases for time series analysis in existing, but also new business models.
Simply by the fact, that it is of huge benefit for a business to know when something is going wrong, or to better
predict whats happening next.
Our solution is standing on the shouldes of giants. We try to build the least possible infrastructure and coding as possible and concentrate on the value add part for our customers. This is why we rely heavily on cloud and open source technologies.
At it's core AnoFox is a set of time series analysis algorithms, which without training data. This is called unsupervised learning. Further they adapt to new data while running, a characteristics of online algorithms.
Enterprise demands information quicker. IoT even requires near realtime reactions. AnoFox support streaming as well as batch operation, and can data process as it arrives.
To be robust, AnoFox uses different algorithms, which different strength and weaknesses. The results of this algorithms are automatically combined, leading to robust automatic analysis for a lot of scenarios.
The digital native companies like Google, Facebook or Netflix provide a large part of their data plattform as open source software, free of charge. This 'best of breed' software is supported by a large global community and makes an ideal foundation for AnoFox.
Data Science means experimentation. Therefore we chose the cloud as an ideal technology platform for our software building blocks, with zero initial investment. We support the major public cloud Amazon AWS, Google GCP and Microsoft Azure. Connection to the data services on these plattforms is easily possible.