Anomaly Detection powered by AI
Machine learning helps detecting anomalies in your data without the need of labels!
What we do
Anomalies or outliers are rare and unusual patterns which can occur in any data stream. craftworks builds AI powered outlier detection systems that use machine learning techniques like deep learning for anomaly detection.
In some cases, identifying such anomalies can form an approach for tackling core business problems, as in the examples of fraud detection (Banking/Fintech) and intrusion detection (Infosec).
In the manufacturing industry, anomaly detection is often a useful tool for enhancement of quality control and machine maintenance practices. For instance, in the context of quality monitoring, an outlier detection system can be used to:
• Assist the personnel in discovering or preventing quality issues
early in the production process
• Simplify and reduce the amount of human inspection needed to
achieve the necessary quality standards
• Identify candidate causes for quality issues
What we offer
Building an anomaly detection system requires a choice of a similarity metric - a measure of likeness between two data points which enables the algorithm to pinpoint the unusual observations. In rare cases, the choice of the similarity metric follows from past experience with similar problems or data sets, and a useful anomaly detection approach can be implemented with minimal labelled data and domain knowledge. More often, however, development of the detection approach requires careful consideration of the relationship between data generation and the monitored process, for which labelled data and domain understanding are required.
At craftworks we provide our clients with detailed guidance on the applicability of anomaly detection solutions to their use cases. Where necessary we carry out in-depth feasibility studies and build prototype anomaly detection algorithms to ensure that the potential to deliver real business value is present in the early stages of the project. During deployment of the full scale solution, we work closely with our clients towards a seamless integration of our system in their IT infrastructure.
How you can work with us
A selection of our Industrial AI clients
With craftworks it was simple and uncomplicated. After an initial conversation, we could start a project within a short amount of time. While the data scientists at craftworks supported us in drawing valuable conclusions from our data, the close collaboration together with their domain experts made it possible to have fast results within a month.
Wien Energie, Head of Innovation and Strategic Projects