Treant Technologies @TreantTech

What’s the single most useful thing an intelligent home camera system could do for you? Better alerts? Searchable footage? Fewer false positives? Something else? Curious what actually matters most.

For my first post, I’m sharing a letter @NVIDIA signed on why open models matter. AI will transform every industry, power every company, and be built by every country. Open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty. The world needs both frontier closed models and frontier open models. images.nvidia.com/pdf/Open-Weigh…

keyboard_arrow_left Previous keyboard_arrow_right Next

@acetomato1 @SteveSmith56721 A lot of the risk people mean is vendor AI in the cloud. Footage and AI on your own hardware stay under your control, not shipped off to someone else.

Home security that stays home. Your AI. Your data. Built for your property, not a citywide camera network. Sign up and take our 2-minute survey 👇

Tired of your phone lighting up for every small movement? Potoo™ (poh-TOO) uses on-device AI to understand what’s actually happening: people, packages, and vehicles. It filters the noise and only notifies you when something matters. Video and AI stay private on your device.

It's good to see the practical application of Ultralytics YOLO26 to real world issues

Ultralytics @ultralytics

4 days ago

Detect pipe damage with Ultralytics YOLO26! 🔧 Identify cracks, corrosion, and surface defects in pipes in real time to support preventive maintenance, infrastructure safety, and industrial inspection workflows. Get started ➡️ bit.ly/4pRp9N6 #Ultralytics #YOLO26

0 22 158 10K 82

A lot of people see any AI camera and immediately think of Flock or public tracking networks. We are the opposite model. Potoo™ (poh-TOO) runs on your device, on your property. Video and AI stay local. No shared database, no agency access, no cloud uploads of your footage. Your cameras. Your data. Your control.

@SteveSmith56721 Hardwire what you can. Ethernet to the box and cameras is the safer path, and we prefer that.

@SteveSmith56721 On this device it does. Footage and AI stay local; you own the data.

@Unclebarney1966 Hardwire what you can. Ethernet to the box and cameras is the safer path, and we prefer that.

Know exactly when someone pulls into your driveway. Potoo™ uses on-device AI to detect and track vehicles in real time, with zero cloud uploads. Tell us what matters to you → Sign up and take our 2-minute survey 👇

@Unclebarney1966 We feel the same. Don't trust the marketing: unplug the internet and the magic still works. Footage and AI stay local on your hardware. Your data, kept to yourself. That's the opposite of Flock.

This is a very interesting article that directly touches on fundamental aspects of the product development we are undertaking at the moment. For any local, on-device AI system the need to provide a self-informing context is paramount to maintaining the relevance of the system. The initial context needs to evolve as the household, the 'organization', develops and acts upon an understanding of how they want to utilize the system capabilities. How far to go down the path to making the graph part of how the AI thinks, not just where the data lives, is definitely an early decision that needs to be made with the understanding that the decision can have an impact on the ongoing relevance of the system as the environment evolves.

The Year of the Graph @TheYotg

a month ago

Layers of Meaning: Context Graphs, Graph Memory, and Ontologies for AI What does it take to build an AI system that actually knows what it means? Not what it says. Not what it retrieves. What it means. Everyone is building context layers and ontologies for AI now. Context

3 2 14 1K 10

Good question. Any license plate read would depend heavily on camera quality, angle, lighting, and stream quality, and that is not our starting focus. We are leaning on other methods to tell if a vehicle looks familiar, which should work better with less capable cameras than trying to read plates. Recognition details are still under active development.

Thanks for the questions @Johnboy7141 1. Video processing and storage are local on your device. We do not store your video in our cloud, and we are not building a shared database for agencies or advertisers to mine. The point is your property, your footage, on your device. 2. The AI engine runs on-device (SylvanAi), on the equipment at your place, not a remote cloud AI service. Learning and updates stay fully local under your control.

@1sevensix WHY?

@abovemilton Yeah. Difference is network and access: Flock is public-road dragnet infrastructure shared with agencies. We're the opposite model: your property, your device, video stays local.

@5thGenKansan Respect. If that setup still covers you, no reason to switch. We're mainly for people who want better alerts without sending video off-site.

@pinoy1954 Good question. Driveway vehicle alerts are just one example. Potoo is private on-device AI for your own cameras and property, so you can focus on what matters at your place even if you do not have a driveway. Video and AI stay local on the device.

Detect power line tower components with @ultralytics YOLO26! Identify insulator strings, crossarms, and crossarm turrets in real time to support power grid inspections, predictive maintenance, and safer monitoring of energy infrastructure. #AI #Energy #MachineLearning

@HotDiggetyDam Hear you on gear that dies after a month. We're early, but long-term reliability on your property is exactly what we're building for, not disposable junk.

@SBJDFW Sounds good. We will share a ballpark when we have one solid enough to put out.