Gracenote, the content data business unit of Nielsen, announced the launch of a new product enabling TV platforms to deliver next-generation user experiences leveraging powerful Large Language Models (LLMs) and gold-standard Gracenote entertainment data. The Gracenote Model Context Protocol (MCP) Server connects LLMs to Gracenote’s continually updated knowledge base and validates, corrects and enriches responses to entertainment queries in real-time. This ensures platforms can instantly return the most accurate and relevant information to users based on their content search inputs and discovery preferences.
"Gracenote content data and IDs have long served as the source of truth for the global entertainment industry,” said Tyler Bell, SVP, Product at Gracenote. “With the launch of the Gracenote Video MCP Server, we're introducing a new wave of offerings leveraging both our data and technology to help platforms solve big problems around engineering and harmonization. We look forward to helping our customers deliver elevated user experiences that position them for success against key engagement and monetization objectives in the AI age.”
Gracenote will initially roll out its Video MCP Server to enable CTV platforms and apps to benefit from the advantages of LLM-driven inference for search and discovery while mitigating their limitations. Dynamically connecting to any LLM, the product verifies and grounds responses in editorially-vetted Gracenote entertainment data. This enables advanced conversational search, highly personalized recommendations and compelling discovery journeys leveraging the world’s most comprehensive collection of human-verified TV, movie and sports data.
With the Gracenote Video MCP Server, TV platforms can answer queries, recommend programming or drive tune-in based on a vast range of parameters. In addition to powering advanced entertainment search and discovery, the product enriches LLM responses with related program imagery, availability information and standardized Gracenote content identifiers. This unlocks data harmonization across different sources and allows linkage to related content such as reviews, trailers and ratings making query responses more useful, authoritative and rich.
TV platforms today are enthusiastic about the power of AI to improve search and discovery performance in order to deepen user engagement and increase monetization. In particular, LLMs, a type of generative AI, provide the ability to create a customized and harmonized content catalog drawing on an unparalleled breadth of training material. Additionally, they enable platforms to offer useful semantic search and the ability to rank and order results against virtually any criteria a viewer or editorial team may be interested in.
But the output of LLM models, even with grounding, can be problematic. Notably, LLMs generate inaccurate but plausible responses to search queries, a phenomenon known as “hallucinating.” Further, LLMs are limited to information included in their training, meaning their inherent knowledge only goes up to a fixed point in time. They also lack access to visual assets tied to structured data like TV shows and movies.
The Gracenote Video MCP Server is the first product from the company’s new suite of AI offerings developed to enable breakthrough content search and discovery capabilities. As consumers increasingly rely on AI and LLMs across all facets of their daily lives, Gracenote’s human-verified data will serve as the source of truth for all entertainment experiences across Video, Sports and Music.