February 26, 2016 | Written by: Denny O’Brien
Share this post:
It’s a fact: IT challenges organizations face every day are increasing, while the staff needed to resolve these issues is decreasing. We see this trend across markets—transportation, finance, broadband and retail. For example, financial options trading halted on the biggest US exchange following an outage caused by software issues. And a leading freight company lost $120 million in revenue because IT was unaware of critical warning messages.
Most of these issues can be traced to the fact that organizations weren’t alerted to emerging issues. But what if they could? That is the power of cognitive operations and systems – they understand the world through sensing and interaction, reason using hypothesis and arguments, and learn from experts and data. The key to cognitive technology like Watson is three-fold:
- Continuous learning – understanding normal behavior and establishing and maintaining thresholds across applications
- Anticipate and adjust –detecting emerging issues across applications and services and proactively alerting and adjusting to changes
- Recommend actions –providing expert advice for corrective actions and greater service assurance.
Machine learning is helping us understand normal operational behavior across applications, infrastructure, and networks, understanding how metrics behave together, identifying problems before you know where to look, detecting service impacts that are not identifiable and assisting with root cause analysis by indicating the most offensive metrics.
Take, for example, a communications agency that has been able to reallocate staff, saving $300,000 annually because of continuous learning capabilities that dynamically set and manage application performance thresholds. As a result, it avoided network outages and improved customer service. They needed to be able to monitor a customer base of 250,000 access lines, 125,000 Internet connections and 30,000 videos. They also need to manage manual thresholds within the networking environment. The solution? Cognitive technology generated alerts automatically when something went wrong, which enabled earlier detection and insights into issues not detected by the normal system. Further, because cognitive technology retains knowledge, the company can easily obtain impact analyses into how the network coped with failure conditions.
Cognitive technology is changing the way the world, humans and business works. We have more control by delegating problems, letting cognitive solve daily issues while humans tackle and mitigate risk. That’s pretty cool.
Learn more about cognitive solutions with IBM IT Operations Analytics.