IoT Anomaly Detection Platform

Modern, data-driven businesses want to move from fire-fighting to proactive support patterns.

The Anomaly Detection Platform provides a configuration-driven, adaptive data pipeline for identifying, classifying, and visualizing anomalies in device telemetry or customer behaviour, for proactive support and maintenance, improved customer experience, and reduced customer churn.

Building on top of endjin's Modern Data Platform and utilizing the full power of Cortana Intelligence Suite, the Anomaly Detection Platform automatically identifies anomalous behaviour in independent events, shows correlations across events to define higher level patterns, and enables data exploration through custom interactive dashboards.

Key Benefits

  • Time series anomaly detection
  • Seasonality adjustments
  • Supports dynamic data schemas
  • Interactive and exploratory visualisations
  • Fully DevOps enabled
  • Built on PaaS services

The challenge for big data driven industries

When you're dealing with massive amounts of telemetry data every day, detailing every user interaction or every system event, how do you spot the unusual activity before it's too late?

The sheer amount of data to process often means that issues that could be identified through manual or ad hoc reporting are already having consequences on customer experience or, even worse, customer churn, before action could be taken.

In addition, identifying patterns and understanding correlations between the seemingly unrelated events is practically impossible due to the sheer size of the data, number of different event types and massive variations in quantities of scale between the different events.

And finally, seasonal variations - at daily, weekly, and monthly aggregations - means that manually identifying behaviour or activity as unusual is subjective and requires in-depth domain and product knowledge.

A turnkey solution to proactive support

Built on top of endjin's Modern Data Platform, the Anomaly Detection Platform is a dynamic, automated, intelligent system that can reliably recognise and ultimately predict patterns of unusual activity across an entire set of telemetry data that could impact on customer experience.

The platform is designed to be maintenance free, so new types of events appearing in the telemetry data are automatically detected and added to the processing logic. Dynamic data schemas are controlled via simple configuration meaning the processing pipeline are adaptive to future data and schema changes, yet can also be easily configured for specific scenarios if necessary.

A series of interactive and dynamic exploratory dashboards are provided, giving on-demand access to the latest data and enabling interrogation of anomalous activity at a macro level, seeing correlations across independent events, and drilling down into the underlying raw data.

The entire platform is DevOps enabled – deployable automatically through a set of declarative templates and scripts that formed part of a continuous integration and deployment process.

Frequently Asked Questions

How do I get going? >
The Anomaly Detection Platform can be purchased as a standalone solution to process your telemetry data. Typically we help you configure your initial data schemas or ingestion pipelines – this helps us ensure your environment is configured for your specific needs and helps us to introduce the processes and knowledge that allow data teams to quickly become self-sufficient. We also provide extensive video labs that walk you through the platform.
I'm a start-up, this looks perfect for my idea, but I don't think I can afford it >
We love working with start ups! We are happy to license IP, consider joint ventures, or provide services for equity. See our Incubator page for more details.
I want to know more, what do I do? >
Chat with us online, pick a time for us to give you a call, or drop us an email at