Pyler taxonomy
Our own categories, used to judge over a billion YouTube videos.
- Controversy
- Hate
- Politics
- Sexual content
- Violence
YAHO understands every scene of your videos and ad creatives, and judges each one against Pyler’s own taxonomy, proven on over a billion videos.
It analyzes what makes up each video, based on its audio, visuals and subtitles.
Categories refined on over a billion videos
Allow, block or review, with the category and reason
Videos and ad creatives, both analyzed scene by scene
A video-level label tells you about the whole video. A scene description tells you what’s there, but not whether it’s OK. YAHO gives every scene both.
Video-level label
Typical brand safety tools
Whole video only
One verdict for the whole video, so you can’t tell which scene to act on.
Scene descriptions
General video AI
No verdict
You can see what’s in each scene, but you still have to decide whether it’s OK.
YAHO scene verdicts
Understand + judge
Review
Disaster, accident (custom), vessel at risk of capsizing detected
open_sea.mp4 · 48:20 · At 01:05, a vessel in distress
Video-level label
Typical brand safety tools
Whole video only
One verdict for the whole video, so you can’t tell which scene to act on.
Scene descriptions
General video AI
No verdict
You can see what’s in each scene, but you still have to decide whether it’s OK.
YAHO scene verdicts
Understand + judge
Review
Disaster, accident (custom), vessel at risk of capsizing detected
YAHO reads what a video shows and its context, and uses the Pyler taxonomy to tell what each scene means. It applies your policies and returns every verdict with its reason.
Our own categories, used to judge over a billion YouTube videos.
A labeled dataset built to the GARM brand safety framework that the ad industry still works from.

Add custom categories and allow or block lists. Lists apply in order: global, local, then organization.
The taxonomy behind YAHO has judged YouTube videos for global brands every day. YAHO takes it down to the scene, for every kind of screen.
pyler on
Long-form shows and channels, scene by scene
Files, URLs and MRSS feeds
Scenes analyzed as they air
Analyzed scene by scene for ad matching and blocking
Solutions put YAHO’s scene understanding and verdicts to work on a specific job.
For media platforms and ad ops teams
YAHO analyzes each creative scene by scene, including its products, mood and message, and returns moderation verdicts with reasons against your policy. Your team approves, fixes or rejects it in your own workflow.
Learn more


For brands and agencies
Describe your brand or the scene you want. YAHO combines scene signals such as place, time of day, activity and mood into moments and shows the scenes behind each one. Scenes with a block verdict are left out. You choose where to run.
Learn moreMoments
Morning · Outdoor · Running
6hr 12m total1,240 scenesDecision : Allow
Race · Finish line · Celebration
1hr 55m total410 scenesDecision : Allow
Night · Road · Collision
Excluded38 scenesDecision : Block
For publishers, platforms and ad ops teams
Every scene gets a verdict with the category and reason from the Pyler taxonomy. Block ads next to scenes that break your brand rules, and show buyers the reason behind every scene you sell.
Learn more
For FAST channels, publishers and SSAI platforms
YAHO finds scene boundaries where an ad break won’t cut off dialogue or action, and gives the reason for each. It works best on VOD and FAST, where breaks can be prepared before airing. Your SSAI or playout sets the final cue points.
Learn more

