Templates
Complete use cases, each one an outcome, the namespace configuration that produces it, and the proof. Every template is a diagram that generates its own manifest, so what you read is what gets deployed.
Stills are illustrative. The boxes are authored to show where each template's decisions fire.
Creative DNA for Ad Archives
Search every scene of every ad you have run, and find the ones that keep earning spend
2,306 scenes across 271 source videos
Face indexedScene embeddedclearedFootage Intelligence at Archive Scale
Millions of scenes you can move through instead of search. Footage, ads and editor assets land in three clean collections, and scenes are clustered one partition at a time so every run stays under the sample-size caps.
1.27M scenes in the reference corpus
sceneclearedVideo Moderation Extraction Foundation
Segment a video corpus, extract a keyframe record for every segment, and match brand marks against a reference set you upload. Face and public-figure matching run on a dedicated or customer-hosted deployment. This deploy is the extraction layer a moderation policy sits on, and the page carries a state on every part so you can see which policy pieces are yours to add.
82% of minors caught at a cutoff of 18
face 0.93 · est. 16 to 18 · borderlinereviewContextual video match
An article goes in, a ranked playlist of video comes out, and every match carries the entities, keywords and categories that produced it. Both sides are embedded twice, once by a text model and once by a multimodal one, so nothing is embedded at query time and a clip in another language still matches.
287ms warm retrieval, no reranker
scene segmentstranscriptclearedUX session analysis
Moderated sessions become a corpus you can question. The transcript is joined to what was on screen at that second, so one query reaches every session at once and returns the moment rather than the recording.
7,257 / 319 word-timed words joined to screen states
participantscreen state at that secondclearedRTSP Live Monitoring Wall
Turn always-on RTSP cameras into a searchable monitoring wall where quiet footage is skipped before extraction, kept segments carry gate verdicts, and email alerts fire from retriever results.
4 live NYC DOT cameras at once
Camera A · searchableCamera B · processingclearedSingle-retriever recipes you can instantiate into an existing namespace.