Semantic Film Intelligence
Your film archive,
understood.
CineIQ analyses the semantic content of video — dialogue, narrative structure, visual composition — and automatically extracts the frames, clips, and trailer assets your platform needs.
Capabilities
Beyond metadata. Beyond keywords.
CineIQ doesn't index your archive — it understands it. Three parallel analysis streams build a complete semantic map of every title.
Dialogue & Narrative
Whisper-based transcription with speaker diarisation. LLM story analysis identifies narrative threads, emotional arcs, and key dramatic moments across the full film.
Visual Analysis
Multi-pass frame scoring: shot detection, scene stability, aesthetic quality, face recognition. Finds the frames worth showing — not just any frame.
Audio & Emotion
Music detection, silence mapping, impact moment identification. Emotional tone tracked across the timeline — fear, tension, humour, romance.
Pipeline
One input. Three streams. One semantic map.
Every title is processed through a DAG of independent analysis filters. Results are merged into a single structured timeline — the film's semantic fingerprint.
Output
Assets ready to publish.
From a single video file to a complete set of platform-ready assets — automatically, at scale.
Poster frames
Cinematically strong, aesthetically scored frames — suitable for thumbnails, posters, and platform cards.
Trailer clips
Story-balanced clip selection covering all narrative threads — not just the most dramatic moments.
Rough trailer
An assembled rough cut with timing cues — ready for editorial review and final polish.
Contact
Let's talk.
CineIQ is in active development. If you manage a video archive and want to discuss how semantic film intelligence could work for your platform, reach out.
Built by Sitlab Limited — Registered in England and Wales