Overview
The Gracenote Video MCP Server supports three primary use cases. Each use case addresses specific business needs and comes with working examples.
1. Catalog Harmonization
Harmonize your content catalog with Gracenote metadata through batch processing.
Catalog harmonization merges existing media catalogs with Gracenote metadata. For example, one implementation of catalog enrichment takes a list of titles from a CSV file or database, resolves each one against Gracenote's catalog to retrieve metadata, including TMSIDs, genres, cast, release year, and descriptions, among others. The process then writes the enriched data back to your CSV file or database. The example in this documentation, runs as a command-line tool or scheduled job, making it well suited for backend data pipelines.
This use case is useful for content platforms that need to normalize and standardize their libraries, data engineering teams building ETL pipelines that incorporate Gracenote metadata, and operations teams that run nightly jobs to keep availability and metadata current.
Typical scenarios include enriching a CSV of 10,000 movie titles with TMSIDs and genre classifications, running a scheduled nightly job to refresh streaming availability across your catalog, deduplicating a content database by mapping internal titles to Gracenote's identifiers, and populating a CMS with cast, crew, and description fields pulled from Gracenote databases.
→ Catalog Harmonization Tutorial
2. Search & Discovery
Build natural language search interfaces that combine Gracenote metadata with LLM understanding.
Search and discovery lets users ask questions in plain language and get back ranked, relevant results enriched with Gracenote metadata including TMSId, cast, genres, descriptions, and streaming availability. The LLM interprets the intent behind the query, translates it into the appropriate MCP tool calls. It then assembles a complex multi-criteria response not possible with a traditional keyword search.
This use case fits any product where users need to find content interactively. Streaming platforms can add natural language search to their apps. Voice assistants can resolve spoken queries like "find me a comedy from the 90s". Chatbots can offer conversational content discovery. Because results are returned in real time, the experience feels like a conversation rather than a database query.
Example scenarios include a user asking where to watch a specific show, searching for the best sci-fi movies of all time, looking for thrillers starring a particular actor, or filtering horror movies from a specific decade that are available on a given streaming service.
3. Personalized Recommendations
Build personalized content recommendation engines that analyze viewing history and generate tailored suggestions.
Personalized recommendations combine a user’s demographic and behavioral data, such as a user's watch history, with Gracenote data to generate highly tailored suggestions. The LLM analyzes demographic indicators and patterns in the history (preferred genres, favorite actors, thematic preferences). It then resolves each title against Gracenote's catalog to understand its attributes, then finds new titles that match those patterns. It can return suggestions with explanations for why each title was recommended. Additionally, results can be filtered based on streaming service so recommendations are actionable by the user.
This use case powers the "because you watched" experience on streaming platforms. It also supports personalized homepage carousels in content apps and curated editorial recommendation lists. It can also support backend recommendation APIs that serve suggestions to any frontend. Personalization can work as both a real-time interactive experience (a user asking for suggestions in a chat), or as a batch process (generating recommendation lists for an entire user base).
Typical scenarios include a user who watched Breaking Bad, The Wire, and Ozark asking what to watch next. A streaming service generating "similar to what you've been watching" suggestions filtered to its own catalog. A viewer who loved True Detective season 1 looking for shows with the same tone and style. A platform generating personalized watchlists based on user ratings.
→ Personalized Recommendations Tutorial