Story & World Intelligence

Film search, AI characters, and interactive stories.

A transcript misses much of what happens on screen. Search and character experiences also need to track who appears, what happens, and how scenes connect, with references back to the footage.

For platforms, studios, and IP holders building search, character experiences, or interactive stories from their films and worlds.

A visitor holds a small token up to a wall-sized screen where a glowing, star-filled character reaches out a hand, while a small white robot on a desk beside her gestures toward the same screen.
One story, two bodies: a character on a screen and a small robot, driven by the same engine.

Worked example

Film search and characters for Escape.ai

On Escape.ai’s film platform, Story Engine watches each film once. The pipeline splits footage into shots and reads out time-coded captions, entities, events, dialogue, and narrative acts. Character grounding ties every appearance to one identity across the film.

Those facts land in a typed story graph of scenes, beats, arcs, conflicts, and themes, with provenance on every node, so an enhancement pass can be reverted without losing what came before it.

Search, character conversations, and interactive stories all read from the same graph. Answers include references to relevant footage, and agents can query and author the graph through a command-line tool and an MCP server.

Diagram: footage is split into shots, analyzed into time-coded captions, entities, events, and dialogue, tied to one identity per character, and enriched with beats, arcs, conflicts, and themes. All of it lands in a typed story graph with provenance, which feeds semantic search with citations, character agents with voice, agent orchestration, and interactive stories. Authoring tools for agents sit below.
From footage to a story graph, and what reads from it. Drawn by us; no film footage or screens.

Read the case study: Story Intelligence for Escape.ai

How it works

  • Story Engine

    keeps a character to one identity across a film, records where every fact in the graph came from, and answers with time codes.

  • DASP

    is the open protocol designed to carry story decisions to engines and rooms as typed messages, so a scene can drive a screen, a headset, or a venue.

  • Spatial OS

    is designed to route a story’s cues to the devices in a room, so a projected character and a robot host follow the same scene.

What we can build

  • Video intelligence

    Shots, captions, entities, events, and dialogue indexed from the footage itself, each tied to its moment.

  • Semantic search

    Questions answered with time-coded citations into the film.

  • Story graphs

    A typed graph of scenes, beats, arcs, conflicts, and themes, with provenance on every node.

  • Character intelligence

    One identity per character across a film, with persona and voice for conversation.

  • World context

    Knowledge graphs that span several works and keep a world’s facts consistent.

  • Narrative-driven agents

    Retrieval agents, multi-character scenes, and interactive stories that read from the graph and stay grounded in it.

Projects

Projects in this area.

All projects

Where to start

Start with one film or series

We index one film or one series into a searchable story graph and put a question-answering agent over it, so your team can judge the answers against the footage.

What you get
A searchable story graph for one property, an agent that answers from it with references to the footage, and a short report on the results and gaps.
What you bring
  • The footage, with the right to process it
  • The questions your team or audience asks about it
  • One person to review the answers

Get in touch

Tell us what you have in mind.

Describe it in a few sentences. You get a draft brief to review before anything is sent.