Visual research for editors
Turn a script into
a visual plan.
SLAYERS reads your narration, works out what each moment needs to show, searches real sources for it, ranks what comes back, and lays the whole thing out as a board you can work from.
Runs the real pipeline against live sources
Narration
- The software industry is undergoing a massive shift as AI coding agents emerge.
- Developers no longer spend hours writing boilerplate code manually.
- Modern IDE interfaces now feature AI pair-programmers inside the editor window.
- Small engineering teams can build complex software in a fraction of the time.
Visual intent
Product UI
Search query
ide ai pair programmer editor
Timecode
00:16 – 00:24
Ranked candidates
- 92
Recommended
Wikimedia Commons
- 74
Alternative
Pexels
- 43
Flagged
Brand source
Illustration of the board structure — run a script to see real results
What happens to your script
Five stages, one pass
- 01
Script
Paste narration, a transcript, or an article.
- 02
Understand
Split into scenes with timecodes and read what each beat is about.
- 03
Find
Query Wikimedia, Pexels, Unsplash and brand sources concurrently.
- 04
Rank
Score every candidate 0–100 on four explainable factors.
- 05
Build
Assemble a scene-by-scene board and export it as CSV or JSON.
What the engine actually does
- Scene segmentation
- Narration is broken into ordered scenes, each with a start and end timecode.
- Visual intent
- Each scene is classified into one of 16 intents — product UI, logo, diagram, location, data visualisation and so on — which decides how it is searched for.
- Concurrent sourcing
- Providers are queried in parallel and isolated: one failing source degrades the result set, it does not fail the run.
- Explainable scoring
- Query match, visual type, source quality and scene context each contribute a fixed share of the 0–100 score. The breakdown is shown on every card.
- Licence capture
- Whatever the source publishes about licensing is recorded and displayed verbatim, with a link back to the source page.
- Export
- The whole board — every scene, requirement and discovered asset — exports to CSV or JSON.