The full Owl-NVR system, end to end.
How the pieces fit together, what the software actually does, and the tech behind every claim. Start with the architecture diagram, then jump to whichever capability you’re evaluating — AI detection, LLM threat reasoning, face recognition, semantic search, health telemetry, or the full software spec.
How all the pieces fit together.
Cameras feed the recorder over PoE. The recorder runs AI locally and serves web, PC, and mobile clients. Your video never has to leave your network.
Any ONVIF / RTSP camera. Bullet, dome, indoor, outdoor — up to 48 per appliance.
One cable per camera carries both power and data. Plug-and-play.
GPU-accelerated AI, multi-channel recording, weeks of on-disk retention.
Live view, smart alerts, semantic search — from any device on your network or via secure remote tunnel.
Everything a modern surveillance system should do.
One appliance, one screen, your whole property — quietly watched by an AI that knows the difference between a leaf and a stranger.
Every camera watched in real time.
AI runs on every stream as it comes in, drawing labels right on the live view as people, vehicles, pets, and packages appear. Sensitivity, watched zones, and which classes to alert on are tunable per camera — so leaves in the wind and passing cars don't generate noise.
- Real-time labels overlaid on every camera stream
- Per-camera sensitivity, watched zones, and ignored areas
- Tracks objects across frames — not single-frame guesses
- Knows people, vehicles, animals, packages, plus custom classes
Plain-English summaries for every event.
Detection sees what is in the frame. Owl-NVR explains what it means. For every event, the AI writes a short scene summary, scores how concerning it looks, and lists what stood out — so you stop wading through 200 motion alerts a day.
- Three-level threat scoring: normal · suspicious · immediate, with a reason
- Intent reasoning — “testing the door”, “loitering near gate”
- Per-camera context (“this is a parking lot”, “this is a back gate”)
- Daily and weekly recaps on demand
- Choose on-device AI for privacy, or a cloud provider for more horsepower
Search your footage in plain English.
Every tracked object gets a short LLM-written description focused on intent and behavior. Those descriptions are indexed alongside object metadata, so queries like “person approaching the front door at night” or “white SUV circling the lot” resolve to actual clips.
- Embedding + keyword hybrid search
- Filters compose with semantic query
- Time-of-day, camera, and zone constraints
- Saved searches (coming soon)
Production-grade telemetry, baked in.
Owl-NVR exposes GPU utilization, inference latency, FPS per stream, RAM, storage, and camera online status on the dashboard. Spot a struggling stream before users do, and trace it back to the camera, model, or hardware.
- CPU / RAM / GPU utilization in real time
- Inference ms + FPS per camera
- Storage runway and retention projections
- Status thresholds: Normal / Warning / Critical
Your video. Your hardware. Your call.
Owl-NVR is engineered for on-prem first. Cloud features — if any — are explicit opt-ins, and most users never need them.
Runs entirely on the appliance. No outbound calls required when using local Ollama and local face embeddings.
Per-camera retention policies. Recordings live on the disks in the appliance, encrypted at rest if you enable it.
Docker, systemd, kiosk-mode display, and an upcoming WSS tunnel for remote access without exposing ports.
What the system can do.
The short, non-jargon version of what Owl-NVR includes. For the recorder build, cameras, and PoE switch, see the Hardware page. Engineering documentation is available on request.
The questions everyone asks.
Not unless you opt in. By default, every frame, every face, and every AI decision happens on the appliance in your building. If you choose a cloud AI provider for extra horsepower, only the frames sent for analysis are uploaded — and that's a setting you control.
Any RTSP or ONVIF-compatible IP camera. H.264 and H.265 are both supported. We've tested with Hikvision, Dahua, Reolink, Amcrest, Axis, and many others.
Strongly recommended. The appliance ships with an NVIDIA accelerator for both detection and (optionally) the local LLM. CPU-only operation is possible for small deployments but will limit framerate and LLM responsiveness.
Most camera systems are recorders — they save video and ping you on every speck of motion. Owl-NVR understands what it's looking at: it tells the difference between a delivery and a stranger, scores how concerning each event is, lets you search clips in plain English, and runs entirely on your own hardware with no monthly fee.
Yes. Owl-NVR can dial out to a broker over WSS and accept reverse-tunneled requests from your phone or laptop. Optional, opt-in, and works behind NAT.
Not today. The product is intentionally appliance-first. If a managed offering ships later, it will be optional — the self-hosted appliance will always remain the default.
See it running on your own cameras.
Book a 20-minute live walkthrough, or send us your property details and we’ll come back with a tailored hardware quote. One business day, from a human.