Vision intelligence,
delivered from the cloud.
CieloVision continuously captures faces and bodies, runs on-device landmarks, and asks Gemini Flash for structured phenotypic analysis every minute. All wired into a glassmorphic dashboard.
01 · Capture
MediaPipe on-device
Face + body landmarks, gaze vector and pose, drawn directly over the live video at 60fps.
02 · Analyze
Gemini 2.5 Flash
Structured JSON phenotype every 60s: emotion, demographics, posture, clothing, action.
03 · Visualize
Holo dashboard
Emotional timeline, demographic pies, clothing breakdown, action cloud, gaze scatter.