Nurby runs local vision models on your camera feeds and lets you write rules in plain language, so you only hear about what matters. Nothing gets uploaded to the cloud, and you do not need any API keys. Your footage stays on your own network.
Watch over babies, pets, and elders without streaming their lives to a stranger's cloud. Nurby whispers when something matters and stays quiet the rest of the time.
Nurby secures a home or a business out of the box. Guardian adds a privacy-first view for places that watch over people.
Families follow one person in a facility and nobody else. You get a note when they arrive safely and a check on who picked them up, with everyone else blurred.
Each parent sees their own child. A push on arrival, pickup checked against an approved list.
A calm status that answers the daily call, without watching the cameras all day.
Nurby watches for gentle chest movement and logs a quiet note when it sees it. It only raises an alarm when something is actually wrong.
Face recognition identifies trusted visitors and logs when they arrive, so you can check instead of wondering.
Draw a loiter zone on the feed. Nurby timestamps anyone standing inside it longer than you want.
YOLO detection finds parcels the moment they land. Nurby files a clip with label, confidence, and timestamp.
Unknown faces that are not on your allow list trigger your rules. Nurby tags the clip and sends you an email, and it groups repeat visitors into clusters over time.
If a camera can see it, Nurby can react to it. Sound a siren when a bear shows up, or quietly post sheep counts to your farm dashboard over a webhook. There is no limit on how many rules you set up.
When a bear is detected in the backyard, sound the yard siren for 20 seconds.
Pick a trigger, set your conditions, and chain the actions you want. Nurby writes the finished rule back to you as a sentence, so you can read exactly what it will do.
When a person is detected on Front Door after 10 pm, send an email and record a 30 second clip.
17 models from yolov8n through yolo-world and the 600-class OIV7 family. Labels are sourced directly from the active weights.
Swap providers per camera or per rule. Stay local with Ollama if you have a capable GPU, or route to a cloud provider when you need accuracy without the hardware.
Nurby's API Call action speaks plain authenticated HTTP with templated URLs, headers, and bodies, so a rule can write into whatever your site actually runs. No certified-partner list to wait for, and no vendor lock-in on our side.
Keyless entry, door controllers, and credential readers. Nurby posts the body count that the badge event alone cannot give you.
Every site is odd in its own way. The lighting, the angle, the one forklift that always parks in shot. A camera product that cannot absorb your corrections makes you tune it forever. Here is where Nurby is on closing that loop, honestly labelled.
A rule can run a vision model as a second opinion before it fires, so the borderline calls get looked at twice instead of paging the shift lead.
Every fired event carries a right/wrong control. Correcting one is the whole interaction, with no labelling tool, dataset, or ML team involved.
Nurby reads back your corrections and proposes the change: raise this rule's confidence to 0.86 and 28 of the 31 you dismissed would have stayed quiet. You approve it, it does not apply itself.
Your corrections become examples attached to that rule's prompt, so the model learns what a mis-slotted pallet looks like in your aisle, not in a benchmark.
Type what you remember and Nurby finds the clip. You do not have to scrub through 23 hours of footage across five cameras to spot the package, the visitor, or the dog in the kitchen. The daily digest hands you the morning's highlights in one glance.
The dog was in the kitchen three times last night. Longest visit was 22 seconds at 11:42 pm near the fridge.
Ollama plus YOLO runs on your hardware. Nothing leaves unless you wire it to.
Toggle blur on specific family members in shared timelines. Keep kids private in exports.
NudeNet blurs sensitive frames. Audio triggers fire only on the sounds you pick.
Every rule, event, and action is addressable. Query your cameras from Claude Desktop via MCP, fire webhooks when rules trigger, or wire Nurby into your home stack with structured JSON.
{
"mcpServers": {
"nurby": {
"command": "python",
"args": ["-m", "services.mcp.server"],
"env": {
"NURBY_MCP_TOKEN": "<your-jwt>",
"DATABASE_URL": "postgresql+asyncpg://..."
}
}
}
}
// "who was at the door today?"
// → answered from your own cameras{
"name": "Late-night front door",
"trigger_pattern": {
"type": "object_detected",
"label": "person"
},
"conditions": {
"camera_id": "front_door",
"time_after": "22:00",
"min_confidence": 0.7
},
"actions": [
{"type": "verify"},
{"type": "telegram"},
{"type": "email", "to": "[email protected]"}
],
"cooldown_seconds": 300
}{
"type": "api_call",
"url": "https://ha.local/api/services/light/turn_on",
"method": "POST",
"headers": {
"Authorization": "Bearer {{ha_token}}"
},
"payload_template": {
"entity_id": "light.porch",
"brightness": 255
},
"secret": "hmac-key-for-signature"
}
// triggers fire → porch light turns onRun Nurby on your own hardware with a single compose file. The full backend is open source on GitHub.
Apache-2.0, self-host, no credit card