VO’s AI-Driven Solutions


Categories Artificial intelligence AI - Machine learning ML

Competition is fierce in the video streaming industry. Pay-TV operators face growing opposition, monetization challenges, and ongoing content security concerns. To attract and grow an audience base, they must offer high-quality, personalized content and targeted advertising, including FAST channels. At the same time, they need to keep their content safe from piracy to protect their revenue streams. Viaccess-Orca is helping pay-TV operators tackle these critical issues by leveraging advanced AI technologies across its comprehensive technology stack. VO’s solutions are based on large language models (LLMs), a type of AI that can understand and generate natural language.

LLMs are built on a deep learning model that analyzes a vast amount of text — millions, if not billions, of words — to effectively predict what word will come next. LLMs also have multi-modal ability and are adept at manipulating and generating images, video, and audio. Fine-tuned “small” LLMs designed for specific tasks, such as content recommendation, are proven to outperform large models. Powered by LLMs, VO’s solutions are a game changer for pay-TV operators, enabling them to optimize audience segmentation in Targeted TV Advertising, improve content recommendations, predict QoS incidents, measure QoE, enhance video playback, and quickly detect video piracy. Leveraging AI to Boost Personalization and Monetization AI is at the heart of VO’s Targeted TV Advertising solution, which features a full-stack programmatic infrastructure and innovative TV data segmentation tools for translating viewers’ usage data into highly valuable segments for advertisers. Driven by AI analytics, VO’s solution enables operators to accurately segment audiences based on household composition, life moment events, demographics, and more. With VO’s solution, operators can leverage usage data to create personalized recommendations proven to drive higher engagement and increase viewing time. As a result, there are more opportunities to display ads and a higher revenue potential.

The same usage data is leveraged to make a unique set of profiles and household segmentation, improving campaign targeting by matching every ad with the desired audience. Improving Service Quality With AI VO’s QoX suite is a perfect example of how VO is elevating video streaming experiences using AI. QoX offers an innovative predictive monitoring service that includes state-of-the-art data engineering technologies and workflows — as well as advanced AI and data science capabilities. Trained on data collected from mission-critical TV platform components, and accumulated through VO’s vast deployment footprint, the predictive monitoring service identifies weak signals, giving operators advance warning about possible service incidents.

The service also provides causality determination by allowing operators to pinpoint the affected TV platform component — meaning any issues negatively impacting the service can be identified and resolved quickly — making this an indispensable tool. With VO QoX suite, operators gain an extensive understanding of streaming quality as the suite provides more than 120 data points for each playback session in real time, including video start failure, time taken to start the playback, rebuffering, playback error, stream quality, application version, the CDN being used, video resolution, zapping time, network bandwidth compared with player bandwidth, and more. In addition, the QoX suite offers seamless integration with complex monitoring systems and datamining solutions, providing operators with access to real-time raw data and pre-processed graphs. VO’s QoX suite is also compliant with the General Data Protection Regulation, which is a legal requirement in Europe and sets a legal framework for handling personal information. Augmenting Content Security With AI One of the most recent threats to video streaming is CDN leeching — also known as vampire services — which involves using a variety of different attacks to allow an unauthenticated user to acquire content from a CDN.   Viaccess-Orca is tackling this issue with powerful new AI and ML capabilities on its Anti-Piracy Center. Thanks to AI/ML technology, operators can prevent piracy and cybersecurity issues across the entire content distribution chain — from the service delivery platform to the end-user device.

This is particularly crucial for protecting live sports content.  AI and ML empower operators by enabling them to correlate data between the CDN structure, the Service Delivery Platform, and the DRM licenses to determine if there are any unusual patterns of behavior that warrant further investigation. For instance, the solution looks for an expected number of sessions token for the amount of bandwidth being consumed. When these values do not match, VO’s team investigates further. If CDN leeching does exist, the solution is able to shut down illegal sessions.  Transforming Video Streaming With AI VO deserves to win the Digital Media World Award for its extensive and innovative use of AI across the technology stack. Leveraging AI — LLMs in particular — VO’s solutions enable operators to offer better service quality and personalization, increase video streaming monetization, and reduce piracy, thus helping their business thrive despite challenges facing the industry today.