EX.CO Introduces Advanced Video Recommendation Engine Utilizing Large Language Models

EX.CO, a leading platform in video publishing that drives successful strategies for top media groups, has unveiled an advanced recommendation engine for digital publishers. This innovative engine, powered by large language models (LLM), provides real-time, relevant video recommendations from a publisher’s content library. It streamlines video integration across websites without the need for creating tailored content for each article or manually pairing articles with videos.

Autovia, the UK’s premier automotive content and commerce company behind brands like Auto Express, Carbuyer, and evo, is among the first to adopt EX.CO’s enhanced recommendation engine across its network of websites.

“Context and relevance are essential for delivering an exceptional user experience,” commented Ciaran Scarry, advertising director at Autovia. “The contextual recommendation engine from EX.CO improves our user experience by aligning video recommendations with the content of each page. This adjustment, though subtle, significantly enhances engagement by providing relevant content that captivates our readers during their car-buying journey.”

Also Read: Tavus Unveils Fastest AI Digital Twin for Live Video Conversations

The LLM-based engine processes text, assesses similarities between articles and available video content, and ranks the results to deliver prompt and high-quality recommendations. For publishers needing additional video content, the engine can access EX.CO’s extensive content marketplace, which features thousands of premium videos across various verticals.

“Today’s audiences demand content that resonates with their interests, which presents a challenge for publishers managing extensive video libraries,” said Tom Pachys, co-founder and CEO of EX.CO. “Through detailed discussions with publishers and our own research, we found that traditional tagging and taxonomy methods were inadequate. By incorporating LLM capabilities with machine learning optimization, we developed a next-generation recommendation engine that has already exceeded the performance of our previous best practices.”

The engine delivers video recommendations rapidly, enhancing audience engagement and retention, and potentially improving key performance indicators such as revenue, brand loyalty, and subscription rates.

EX.CO’s updated recommendation engine is now operational with EX.CO partners, achieving an 80% relevancy match rate and resulting in four times higher engagement rates compared to industry standards. Additionally, negative interactions with the video player have dropped by 30-40%.

The technology currently refines video recommendations based on factors like media category, title, recency, sentiment, keywords, and length. EX.CO plans to enhance this offering by incorporating ChatGPT-like functionality, enabling publishers to fine-tune video recommendations using specific prompts. This will further optimize the engine’s ability to deliver highly relevant content tailored to individual sites, sections, and articles.