Skip to main content

eoMind Streaming Analytics

Real-time Subscriber Anomaly Detection


The Automation of Insight

One of the largest challenges in telecoms today is making sense of the huge amounts of data available to management, operations, marketing, engineering and customer care in order to understand the subscribers’ Quality of Experience (QoE). There is simply too much data making it almost impossible to actually identify the issues in a short period of time.

Big data in telecoms is characterized by the large volumes of data (often many billions of event records per day are generated) and the many varieties of that data (the different interfaces, network and application technologies that are monitored). The data can be extracted from network elements, probes, sensors, log files and even from social networks.

For a variety of reasons - mostly a lack of time and resources - this priceless data is often left lying dormant; opportunities for improved service, cost, and customer retention are lost. Allowing this data to sit idle is clearly not an effective use of a CSP’s information asset. Identifying the insights inside the data streams is therefore key to identifying issues that affect your customers and also to respond to threats and opportunities for service delivery.

A continuous process of extraction of information is required to maximize the investment in data sources that CSPs have already made.

Key Target Users

  • Network Operators who need to analyzes huge amount of data coming from their network
  • Operational executives who would benefit from key insights that impact the performance of the network and subscriber experiences
  • Tactical officers who need to identify the issue and possible causes without specialist involvement
  • Financial executives needing analysis in real time -- leveraging a streaming analytics technique that give results in seconds
  • Typical roles interested in these benefits include CIO – Chief Information Officer, CTO – Chief Technology Officer, COO – Chief Operating Officer, VP Engineering, VP Customer Experience, VP Operations
Key Benefits

Reduce MTTR, with fewer impacted subscribers using Augmented Analytics
eoMind uses streaming analytics to detect subscriber-affecting issues. Using patented machine learning algorithms, eoMind automatically detects anomalies in behavior for subscribers, isolates the root cause and then closes the loop using intelligent automation with a customizable suite of next best actions.

It does it faster, with high accuracy, reduces mean time to resolution (MTTR) and, with cases closed faster, significantly reduces knock-on impacts to subscribers. eoMind is real-time Automated Assurance for service providers around the world looking to move to Smarter Operations.

Reduction in MTTR
Proven reductions of 25% in the meantime to resolve issues for customers freeing up valuable expertize to work on critical issues.

Decrease in impacted subscribers
With issues detected faster and closed sooner, overall knock-on impacts to subscribers are reduced with savings on calls to care and fewer escalations.

OpEx efficiency
250 configurable next best actions available for common issues, saving time and money, with the library of actions garnered from customer installations increasing all the time.

Lightweight, data agnostic, impactful
Focusing uniquely on subscriber issues, eoMind is resource-light yet impact-heavy and can be used to target improvements with Corporates, VIPs, particular technologies or device rollouts.

Key Features

Subscriber issue detection using ML
Tailored patented algorithms scan high bandwidth SIRCA (subscriber impacting root cause analysis) data streams looking for subscriber anomalies. Detection of anomalies are changes to subscriber behavior or experience and are surfaced up quickly, using tailored EWMA (exponential weighted moving average) algorithms for immediate remedial actions.

Drill-down and root cause analysis automation
Automated localization of issues, based on subscriber, error, device, and node are made with a weighted root cause analysis assessment highlighting the most likely root cause.

Real-real-time streaming analytics
Detection of anomalies are predicated on anomalous changes to subscriber behavior or experience and are surfaced up quickly, usually within the first minute and alongside drill-down and root cause, for immediate actions.

Next best action
Sophisticated next best action tooling processes cases automatically and, based on operational and business need, tailors a suite of next best remedial actions closing the loop without human input required for detection, triage, and repair.