Video understanding · metadata & automation
Turn every piece of content into usable intelligence.
Video Intelligence watches, listens and reads every frame to generate structured metadata for discovery, programming, compliance and editorial workflows.
Multimodal analysis for live, VOD, news, sports and advertising.
The problem
Discovery is only as good as your metadata — and metadata is expensive to make.
Manual logging is slow and inconsistent. Generic labels miss the people, topics, scenes and moments that make video useful.
Video Intelligence turns audiovisual content into structured, searchable and workflow-ready data.
Understand
It doesn't just generate tags. It understands the content.
Speech, image, context and time combined in one analysis.
Detect scenes, topics, people, places, objects, on-screen text and sentiment with timestamps and confidence.
Multimodal analysis
Combine visual, audio and language signals.
Time-coded output
Every result points back to the exact moment.
Confidence scores
Keep automation explainable and controllable.
Organize
Every kind of content, better organized.
One intelligence layer for entertainment, sports, news, live and ads.
Apply specialized extraction to the formats your catalogue actually contains.
Two operating modes
Analyze a catalogue or understand a live signal.
Batch processing
Submit individual files or entire catalogues, process them asynchronously and retrieve normalized metadata when each job completes.
Files · catalogues · archives
Real-time analysis
Connect a live stream and receive time-coded events, entities and programme boundaries continuously as the content plays.
Live channels · events · monitoring
Enrich
Consistent, detailed metadata — ready to plug into any workflow.
From raw video to controlled taxonomies and rich catalogue records.
Normalize output against your vocabulary and route it to search, recommendations, CMS and archives.
Your taxonomy
Map generated metadata to your editorial language.
Search-ready
Improve retrieval with entities, themes and moments.
Workflow-ready
Push structured output to downstream systems.
Operate
Editorial teams work in the UI. Developers automate the same thing over the API.
Human review and production automation share one platform.
Inspect, correct and approve results in the workspace, or trigger analysis programmatically at scale.
Editorial workspace
Review and refine machine-generated intelligence.
Batch processing
Analyze catalogues and archives at scale.
API automation
Integrate ingestion and delivery end to end.
For engineering teams
A REST API your systems can build on.
Submit content, follow processing and retrieve structured, time-coded intelligence through a stable API.
Movies & Series
moviesAndSeriesFilm and series content tagging. Identifies genre, characters, scenes and mood.
Response fields
genreGenre classification
charactersCharacters identified in the content
scenesScene-by-scene breakdown with timestamps
moodOverall mood and tone
Code example
Full workflow using analysis_type_id: "moviesAndSeries"
# 1. Request upload URL
curl -X POST "https://api.aitagging.jumptechnologies.ai/v1/media" \
-H "X-API-Key: tgfy_live_YOUR_KEY_HERE" \
-H "Content-Type: application/json" \
-d '{"filename": "video.mp4", "content_type": "video/mp4"}'
# 2. Upload video binary to the signed URL
curl -X PUT "<upload_url>" \
-H "Content-Type: video/mp4" \
--data-binary @video.mp4
# 3. Create analysis task
curl -X POST "https://api.aitagging.jumptechnologies.ai/v1/tasks" \
-H "X-API-Key: tgfy_live_YOUR_KEY_HERE" \
-H "Content-Type: application/json" \
-d '{"media_id": "<media_id>", "analysis_type_id": "moviesAndSeries"}'
# 4. Poll task status until "completed"
curl -X GET "https://api.aitagging.jumptechnologies.ai/v1/tasks/<task_id>" \
-H "X-API-Key: tgfy_live_YOUR_KEY_HERE"
# 5. Get results
curl -X GET "https://api.aitagging.jumptechnologies.ai/v1/results/<task_id>" \
-H "X-API-Key: tgfy_live_YOUR_KEY_HERE"Product spec
Video Intelligence
Make your content easier to organize, find, and enjoy.
See Video Intelligence working with your own catalogue, audience and data.
.png?alt=media&token=92e1d872-53df-40a9-a3df-b00043a3bae4)
