Hollywood Metrics joins as data partner
Seventy-four thousand films, $803B in tracked box office, 107 years of cinema — now wired into the registry: track records on profiles, comp context on quotes, and a script-to-casting pipeline across two MCP servers.
A likeness license is a bet on a person. Buyers make that bet better with receipts: what this actor's films actually did, how audiences actually rated them, what campaigns like this one actually cost. Today those receipts come built in — Semblance has partnered with Hollywood Metrics, the film-analytics platform indexing 74,571 films, $803B in box office, and 1.1M ratings across 107 years.
What it powers, from today's preview
- Track records on profiles. Screen talent and estates carry filmography rows — credits, tracked gross, average audience rating, active span — verified against the Hollywood Metrics catalog and attributed wherever they appear. Open Imani Sato or the Whitfield Estate in the catalogue to see it.
- Market context on quotes. Bookings in review show comp ranges from similar campaigns and productions, so a fee is a number with neighbors instead of a guess — visible today in the console's booking cards.
- A two-server agent pipeline. Hollywood Metrics runs its own MCP server (17 tools: search, script grading, outcome forecasting). An AI producer can analyze a screenplay and forecast the package there, then search, quote, and license rights-cleared talent here — the full chain is on the API page.
- Casting by comps. Find the films your project resembles, then find registry talent whose track record fits them — discovery that starts from evidence.
The honest part
Bankability scores have a long, mixed history — indices that flatten careers into a number have hurt working talent before. Our rule for this integration: data appears with its inputs, attributed to its source, as context beside a human decision. Talent see exactly what buyers see. Nothing in the pipeline scores a person; it prices a project.
Integration details and the research behind it live in our files; the pipeline is documented on the API page. All catalogue entries remain fictional specimens until wave 01 lists.