AUTHOR=Molnar Gabor , Ferreira Pires Luís , de Boer Oscar , Kovaleva Vera 

TITLE=An empirical assessment of the use of an algorithm factory for video delivery operations

JOURNAL=Frontiers in Artificial Intelligence

VOLUME=Volume 7 - 2024

YEAR=2024

URL=https://www.frontiersin.org/journals/artificial-intelligence/articles/10.3389/frai.2024.1281110

DOI=10.3389/frai.2024.1281110

ISSN=2624-8212

ABSTRACT=Video service providers are moving from focusing on Quality of Service (QoS) to Quality of Experience (QoE) in their video networks since the users’ demand for high-quality video content is continually growing. By focusing on QoE, video service providers can provide their subscribers with a more personalized and engaging experience, which can help increase viewer satisfaction and retention. This focus shift requires not only a more sophisticated approach to network management and new tools and technologies to measure and optimize QoE in their networks but also a novel approach to video delivery operations.

In the PayTV industry, characterized by rapid technological advancement, the static and unchanging nature of traditional architectural designs is becoming outdated and ineffective. Modern practices to address the operational aspects of these architectures must embrace change and accommodate the ever-evolving needs and preferences of market players. This paper describes the components, interactions, and relationships of an algorithm factory for video delivery operation that assures high QoE for video streaming services. The paper also showcases the results of gradually implementing an algorithm factory in the video industry. 

Using a dataset from 2016 to 2022, we present the case of a European PayTV service provider that achieved improved performance measured by both objective and subjective metrics. Improved performance is demonstrated by an increase in the speed of critical incident resolution by roughly five times and a decrease in the number of critical incidents by 59% while growing the customer base and applying the same amount of labor resources. The case also demonstrates a strong positive relation between the productivity measures of the PayTV operator and their survey-based quality ratings. The paper adds to the existing literature on relationships between operational efficiency, innovation, and subjective quality. The paper further offers empirical evidence from the PayTV industry. The insights provided are expected to benefit both traditional and over-the-top (OTT) video service providers in their quest to stay ahead in the rapidly evolving video industry. It may also translate to other service providers in similar industries committed to supporting high-quality service delivery.