All your questions about CuriX answered.


CuriX prevents system malfunctions or even failures. CuriX significantly reduces downtime costs, eliminates performance problems and automates troubleshooting. In addition, valuable IT staff can be reassigned to value-added activities.

CuriX also prevents image problems that quickly occur as a result of a cyber attack or system failure.

CuriX learns the normal behavior of a system based on the monitored data, e.g. the performance metric of a CPU. If a value deviates from the ‘expected normal value’ at a certain point in time, this deviation is called an anomaly.

The ‘CuriX Cycle’ represents the multi-step analysis heart of CuriX. The main steps are

  • Collection, aggregation and correlation of data e.g. metrics
  • Detection of anomalies
  • Prediction of failures based on “critical combinations” of anomalies
  • Localization of potential failure locations for predicted failure prediction
  • Fault recovery by specifying “heal-advices” up to “automatic self-healing”.

CuriX stores results and process data e.g. the learned baseline. The original data collected remains in the source systems and is not stored redundantly in CuriX.

This depends on the dynamics of the system to be learned. Typically, first statements about anomalies can be made with sufficient quality already after 1-2 weeks of learning. By regularly relearning the system behavior, the analysis results can be continuously improved.

CuriX provides various connectors for external systems and can, for example, dock directly to existing ticketing systems such as JIRA to generate incidents or query the status of existing tickets.

While virus scanners look for known patterns, CuriX also detects unknown and novel malicious patterns. Cyber threats and attacks will become more and more intelligent and internally triggered malfunctions will become more and more complex, therefore we close the gap of the unknown and provide protection against future threats of all kinds by the continuing learning system. However, a virus scanner can be used in addition to CuriX or integrated into the CuriX solution.

CuriX is a NextGen resilience tool built on top of a data lake (typically fed by observability, monitoring, or SIEM tools).

CuriX is able to make predictions about the future based on the metrics and logs of classic tools:

  • make statements about the future (predictions)
  • reduce alarms to significant resilience problems (Noise Reduction) and
  • identify critical time periods or system components.

CuriX also issues a “heal advice” for critical resilience issues, which can be automatically forwarded to a ticket system.

Classical tools usually operate with static thresholds, i.e. limit values which, if exceeded, must trigger an action or a critical system state (including potential system failure) occurs. CuriX can calculate a future state based on actual data. In addition, CuriX is able to recognize dependencies of individual issues by means of correlations and to place them in an overall context – from the origin of a problem (root cause) to the future, negative impact on a system, an application or a service.

Metrics, log files and traces can be evaluated.

The integration tool and command center is the CuriX Dashboard. It is web-based, widgetized (so you can customize your view) and provides both an overall view and deep dive capabilities into the individual components of a system. The basis for the health status of a system is the so-called Health Score. CuriX can also write results to other tools or dashboards.

A Heal Advice is a solution suggested by the system for a problem that will occur. The simplest form of a Heal Advice is a ticket that is automatically forwarded to the responsible instance or to a ticket system.

CuriX not only looks outwards, where cyber attacks potentially come from, but also inwards, where faulty manipulations or malfunctions can trigger major damage. In addition, CuriX also detects previously unknown patterns (so-called unknown unknowns). This comprehensive view as well as the automation makes CuriX an “immune system” for IT systems.

CuriX is available as an on-prem, hybrid or cloud solution, adapts optimally to a user’s topology and roadmap (e.g. Journey to Cloud).

CuriX acts as a dome across the entire infrastructure, centralizing all systems into one. Instead of replacing, we use data tools already in the system to make intelligent predictions and avoid unnecessary or late alarms.

CuriX brings existing systems to a higher level. CuriX not only brings together so-called data silos (bracketing function), but also becomes a resilience and compliance tool for a company thanks to its predictive, automated functionality.

CuriX cures infrastructure – and because CuriX also recognizes unknown, novel patterns accurately (unknown Unknows), CuriX is the solution for X-any problems – therefore CuriX stands for: “Cure infrastructure in X (any environment)”.

CuriX enables a company to manage the success and competitive factor IT systems proactively and data-based. In an increasingly digitalized world, a monitoring resilience tool is a basic requirement to ensure smooth operation at all times. CuriX reduces complexity and brings cost-efficient operation.

  • Reduces complexity (saves time)
  • .

  • Enables data based decision making
  • .

  • Holistic view of the entire IT
  • .

  • Merging static and dynamic data
  • Reducing effort through RCA
  • Acting instead of reacting (no more fire drills!)
  • Avoid system failures
  • Automate processes
  • Employees have more time for value-added activities
  • .

  • Reduce complexity (save time)
  • Holistic view of the entire IT
  • Merging static and dynamic data
  • Reduction of effort through RCA
  • Acting instead of reacting (no more fire drills!)
  • Avoid system failures
  • Automate processes
  • Mitarbeiter haben mehr Zeit für wertschöpfende Tätigkeiten


The following settings must be configured:

  • Cycle length for data collection, aggregation and analysis.
  • Filter for data collection
  • Settings for health score
  • Learning phases for anomaly detection
  • Learning phases for failure prediction
  • Credentials for connectors to peripheral systems (ticket system, CMDB, etc.)
    Dispatching of information and alerts (incl. email)

CuriX is delivered on a HW appliance, the effort for the customer is very low, CuriX takes care of that.

The implementation is standardized, the CuriX deployment guide is used for this.

CuriX already offers connectors for the most important source and output systems by default. These are described in the Deployment Guide. If the customer wishes to connect a specific data source, this is free of charge for the customer if the data source can be included as standard in CuriX.

Interfaces to Datalake tools:

  • Elastic
  • Splunk
  • Qradar
  • MF OpsBridge
  • InfluxDB
  • PRTG
  • Whatsup
  • Zabbix
  • Check_MK
  • Prometheus, Graylog
  • Cloud (Azure, AWS, Google, Alibaba)

Interfaces to ITIL tools (CMDB/ticket system):

  • iTop
  • JIRA (Servicedesk+Insigth)
  • ServiceNow
  • Matrix42
  • Sysaid
  • Valuemation
  • ServiceWare
  • Cherwell
  • Topdesk
  • Ivanti
  • Kyberna

Metadata regarding configuration items, service organizations and busness processes are obtained from the CMDB.

CuriX anomalies, failure predictions, limit violations are intelligently forwarded to the correct service organization.

A PoV usually takes 4-8 weeks, depending on the project scope. The effort for the customer is very low, according to experience max. 5 days. A GoLive after the PoV requires no more effort.


So-called nodes are licensed. The customer can analyze any number of KPIs. This keeps the costs under control.

CuriX is paid for with an annual subscription and is delivered with a CuriX appliance.


Very low, minimal maintenance window is provided for automatic updates.

When new measured values from additional systems are added, the data is automatically taught in.

The data resides in a data lake (typically fed by monitoring, observability, SIEM tools). If no data collector tool is available, CuriX comes with an Elastic Stack (Basic Edition) that serves this purpose.