Alive Analytics
The primary objective in the development of the Analytics capabilities of Alive™ was to remain data agnostic. Data agnosticism has come to mean that the Alive™ Analytics can equally perform their tasks on any type of data regardless of source or context. 
In other words Alive™ can perform analysis on Operating System metrics just as it would on business metrics just as it would on datacenter environmental metrics.
Although far more time consuming and difficult to develop such algorithms, the applicability of Alive™ to the entire enterprise brings unprecedented benefits to our customers. Not only can Alive™ analyze network data, Operating system data, application data and virtual data, it can also analyze Batch data, storage data, and business data. The researchers at the Integrien Advanced Statistics Center (IASC) are dedicated in delivering data agnostic algorithms that can span the entire scope of the enterprise for an unprecedented value offering.
The Alive™ Analytics begins by analyzing the time series behavior of every metric that is fed into the software. This analysis consists of eight algorithms for determination of an appropriate baseline (or dynamic threshold) for each metric. None of these algorithms make any assumptions about the data (such as being normally distributed etc.) as IT type metrics never exhibit behaviors that can be characterized as such. The key insight of our years of research into this field is that no one algorithm can appropriately accommodate all types of metrics that could exist in an enterprise.
Thus the researchers at IASC developed multiple algorithms each with its own unique capabilities that compete with each other for determining the best possible threshold for each metric. Some of the capabilities here include:
- Automatically determining the natural cycles in the data. Whether the cycles determine regular business cycles or exceptions to those cycles (such as holidays, payroll, etc.) there is no user input required.
- Algorithms can also perform base-lining on Boolean type metrics (such as availability, ping, port checks, etc.)
- Special set of algorithms can analyze Batch metrics to automatically determine the periodic rate of the jobs and their normal duration period.
- Automatically detects change in (as opposed to an anomaly) and starts the adjustment process of forming new thresholds based on the newly changed environment.
Once the dynamic thresholds have been established on each metric Alive™ then performs Systems Analysis on the aggregate group behavior which then leads to our Smart Alerts. Alive™ also has the capability to mathematically combining raw metrics to form new metrics that maybe more meaningful to the users (called Supermetrics). The combination of Aggregate Anomaly Analysis along with Supermetrics provides a rich set of Systems data that can provide for Alerts that not only notify you before your end-users are impacted, but they also provide the most probable root cause for quick issue resolution.
List of Patents:
- Self-learning, Performance Management System and Related Methods.
- Methods for the cyclical pattern determination of time series data using a clustering approach.
- Nonparametric method for determination of anomalous event states in complex systems exhibiting non-stationarity.
- Dynamic Thresholding via weighted quantile analysis.
- System and method for generating and using Fingerprints for integrity management.
- System and method for correlating Fingerprints for automated intelligence.
- Dynamic problem determination using aggregate anomaly analysis.
