Videos
What it is
Section titled “What it is”The Videos feature enables video-based discovery. Upload client meeting recordings, product demos, or walkthrough videos and Foundry uses the Twelve Labs integration to extract transcripts, keyframes, and structured findings from video content.
Why it matters
Section titled “Why it matters”Valuable project context often lives in recorded meetings and screen shares. Video analysis captures requirements, decisions, and context that would otherwise be lost in hour-long recordings — surfacing them as structured findings alongside document-based discovery.
Key concepts
Section titled “Key concepts”- Video upload:
VideoUploadPagehandles file selection and upload to storage - Analysis pipeline:
PipelineProgresstracks the multi-stage video analysis workflow - Video detail:
VideoAnalysisDetailPageshows transcripts, extracted keyframes, and findings - Visual discovery gallery:
VisualDiscoveryGallerydisplays extracted keyframes with annotations - Activity feed:
VideoActivityFeedshows events from the analysis pipeline - Analysis completion card:
AnalysisCompletionCardsummarizes results when analysis finishes - Twelve Labs integration: External video intelligence API used for transcript generation and scene analysis
How to use it
Section titled “How to use it”- Navigate to [Program] > Discovery > Videos (or the dedicated Videos route).
- Click Upload Video and select a recording.
- Watch the pipeline progress as the video is processed through analysis stages.
- Once analysis completes, review the transcript and extracted keyframes.
- Browse the Visual Discovery Gallery for annotated frames.
- Review extracted findings alongside document-based discovery findings.
Data model
Section titled “Data model”This feature uses the following tables:
videoAnalysis— Video analysis records with transcript and metadatavideoTranscripts— Extracted transcript segmentsvideoKeyframes— Keyframe images extracted from videovideoFindings— Structured findings extracted from video content