Frequently asked
questions

No. Zenvisia works with existing security and traffic cameras. Whether you have standard IP cameras streaming RTSP, older analog cameras converted via an NVR, or recorded video files, our GPU pipeline processes the visual frames directly into calibrated trajectories.

Zenvisia uses perspective homography matrices to project 2D image coordinates into real-world geographic coordinates (like OpenStreetMap). Every bounding box detected on the camera is converted into a physical position with metric coordinates, allowing time-synchronized tracking from scene entrance to exit.

Anything grounded in the trajectory data: entry/exit counts by direction and time window, peak intervals, loitering detection, individual path playback by ID, and simple comparisons across different time ranges in the same clip.

MP4, MKV, MOV, and WebM containers (H.264/H.265) up to 24 hours per clip. For best precision, use 1080p or 4K footage at 15–60 fps from a fixed, elevated camera (3–15m) with minimal fisheye distortion.

You upload any recorded video clip (up to 24 hours in duration). Our background GPU workers process the footage, extract all pedestrian, vehicle, and bicycle trajectories, and generate an interactive dashboard where you can converse with the AI assistant, inspect peak intervals, and download track reports.

Yes — the Pro Surveillance plan ($149/mo billed annually) adds live camera streams, more monthly GPU hours, real-time alerts, and multi-day historical queries. See the pricing page for the full comparison.

We intentionally limit custom background GPU processing slots to keep accuracy high while we scale the pipeline. Reserve your spot on the waitlist from the upload page and we'll notify you the moment a slot opens, with two free analysis hours included.

Yes. Zenvisia is built strictly around movement and volume analytics. It tracks anonymous sequential numbers (#1, #2, #12) and vector paths. We do not perform facial recognition, license-plate extraction, or biometric identity profiling.

Uploaded clips are stored in encrypted cloud storage scoped to your account. Enterprise customers can opt into sovereign cloud regions or fully air-gapped on-premise installations.