Predicting Service Impairments from Set-top Box Errors in Near Real-Time and What to Do About It (2018)

By Justin Watson, Comcast; Roger Brooks, Andrew Colby, Pankaj Kumar, Anant Malhotra & Mudit Jain, Guavus, Inc.

Getting ahead of subscriber problems is a difficult but powerful way to reduce costs and increase customer satisfaction. This paper describes a proof of concept (POC) trial that harnessed machine learning and Comcast X1 service impairment data to identify at-risk subscribers and risk drivers, and to further indicate next best-actions to take in response to the predicted issues.

Machine learning is well-positioned to address the blind spots of customer support teams. It can be architected to scale to tens of millions of simultaneous event streams and handle real-time, complex predictive analytics. The highly accurate analytics used in this trial enabled us to identify subscribers potentially affected by impairments responsible for generating 36 percent of calls and 46 percent of tickets. Only 5 percent of the device population drove these care events. To enable action, our analytics also identified the associated risk drivers. In another exercise, we predicted a large quantity of true positive tickets per year related to 13 newly clustered ticket classes, with known resolution paths, and associated 57 percent of those tickets with single problem codes. (Note: All tickets referenced in this paper are technical tickets.) Shared among internal stakeholders, these kinds of insights can drive numerous benefits. They can enable service providers to proactively address technical problems of subscribers; reduce the number of calls, tickets and truck rolls as a result; and more quickly resolve impairment events that do arise. The net result is reduced costs and improved customer experience.

By clicking the "Download Paper" button, you are agreeing to our terms and conditions.

Similar Papers

Customer First: CX-Driven Augmented Operations
By Roger Brooks, Ph.D., Pankaj Kumar, Mudit Jain, Megha Vij, Nandit Jain & Andrew Colby, Guavus
2019
Low Complexity Real-time Video Encoding for Soft Set-Top Box Platforms
By Krasimir Kolarov, Feei Chung, William Lynch, Interval Research Corporation
2000
Cost Effective Watermarking In The Set Top Box
By Joseph Oren, Cinea Inc. a Dolby Company
2008
Employing Neural Networks for Improved Root Cause Analysis in Service Provider Clouds
By Helen Zeng, Ph.D. & Robert McIntyre, VMware
2020
Evolving The “Box”: The Smart Set-Top Box
By David Goodwin & Charles Cheevers, ARRIS
2018
Embracing Service Delivery Changes with Machine Learning
By Andrew Sundelin, Guavus, Inc.
2018
Considerations For A Real Time Broadband Delivered Interactive Gaming Service
By Charles H. Jablonski, OnLive Inc.
2010
A PNM System Using Artificial Intelligence, HFC Network Impairment, Atmospheric and Weather Data to Predict HFC Network Degradation and Avert Customer Impact
By Larry Wolcott, Michael O'Dell, Peter Kuykendall, Vishnu Gopal, Jason Woodrich & Nick Pinckernell, Comcast
2018
New Developments In IEEE-1394 Standards For The Cable Set-Top Box
By Mark Eyer, Sony Electronics
2003
Implement Closed-Loop Network Decisioning Now with Big Data Analytics and Fuel Future-State SDN Use Cases Through a Common Platform Deployment
By Brennen Lynch and Anukool Lakhina, Guavus, Inc.
2014
More Results >>