✦ Home Automation × AI Vision

Automating the
physical world
with AI.

Bridging the gap between hardware and software — turning everyday devices into intelligent, context-aware systems using Home Assistant, AI vision, and real-time automation logic.

Flash Sale: Limited Time Only

Get 50% OFF Today

Secure the complete AI Pet Feeder Blueprint & Guide for half the price. This offer expires soon.

Featured Project

AI Pet Feeder
with Vision Confirmation

A fully autonomous feeding system that doesn't just schedule meals — it sees, thinks, and confirms.

🐈‍⬛

The Problem

Standard smart feeders are time-based and blind. They dispense food even when the pet isn't there, overfeed, double-feed, or completely miss the pet's presence. There's no intelligence — just a timer.

🤖

The Solution

Motion triggers a camera snapshot. Google Gemini AI analyzes the image to confirm it's your pet. Only then does the feeder dispense the right portion — then a second AI call confirms the kibble actually dropped. Full loop, zero guessing.

How It Works

📡
Step 1
Motion Detected
Tapo camera cell motion sensor triggers the HA automation
📸
Step 2
Snapshot + AI Analysis
Camera takes a snapshot. Gemini Vision confirms: is it the right pet?
🍽️
Step 3
Smart Dispensing
Correct gram portion dispensed based on time-of-day schedule and daily limit
Step 4
Visual Confirmation
Second AI call confirms kibble is in the bowl. You get a push notification.

What's included in the blueprint

Anti-double-feeding cooldown (configurable, e.g. 130 min)
Daily gram limit tracking — auto-stops when limit is reached
Morning double portions, daytime single portions (time-based logic)
Customizable AI detection prompt — works for any pet type
Smart retry window after false detection (no spam triggers)
Single notification per false alarm (failure tracker toggle)
Works with any Wi-Fi feeder exposed as a number entity in HA
Full Home Assistant Blueprint — 100% configurable via UI, no YAML editing
Tech Stack
🏠 Home Assistant 🤖 Google Gemini AI 📷 Wyze / Tapo Vision 🐍 Python Automations ☁️ Cloudflare Tunnels 📱 Push Notifications

Complete setup guide · Blueprint file · Configuration walkthrough

Full Feature Breakdown

Everything This Automation Does

Click any feature to expand the full details. This is what makes it the smartest pet feeder on the planet.

📦

What's in the DIY Starter Kit

  • The Custom Blueprint: The heavily tested HAiDannys_AI_Pet_Feeder.yaml file, ready to import directly into your Home Assistant.
  • The Ultimate Setup Guide: A step-by-step PDF manual covering helpers, auto-cleanup scripts, AI prompts, and troubleshooting.
🛑

What You Need Before Buying

  • A working Home Assistant hub
  • A Wi-Fi Smart Feeder + Pet Camera already integrated into your HA dashboard
  • An active AI Vision API Key (OpenAI, Anthropic, or Google) connected via the LLM Vision integration

Don't have these set up yet? The VIP package includes a 1-on-1 call where we do it together.

Choose Your Package
$50
DIY Starter Kit

The complete Blueprint YAML + the 7-Step PDF Setup Guide. Perfect for users who already know their way around Home Assistant.

Get Starter Kit →
Most Value
$150
VIP "Done-For-You" Setup

Everything in the Starter Kit, PLUS a 1-on-1 video call with HAiDanny. I will personally log in with you and get the AI automation running perfectly on your machine.

Book VIP Setup →
Toolbox

The Stack Behind the Magic

🧠

AI Vision Analysis

Integrating Gemini Pro Vision to distinguish between pets and humans, ensuring 100% accuracy in triggering actions.

⚙️

Smart Logic

Sophisticated Home Assistant Blueprints with daily limits, anti-double-feed cooldowns, and multi-stage verification.

🔌

Hardware Integration

Connecting Wi-Fi feeders, ONVIF cameras, and smart sensors into a unified ecosystem that actually talks to each other.

🛡️

Secure Connectivity

Remote access via Cloudflare Tunnels and local-first control for high reliability and maximum privacy for your home data.

H
About the Builder

Authentic, adaptive
tech explorer.

I'm HAiDanny — a home automation enthusiast who got tired of "smart" devices that aren't actually smart. So I started building the missing piece: automations that see, think, and respond.

My approach is practical: take consumer hardware, layer in real AI vision, orchestrate it with Home Assistant, and document everything so you can replicate it. No PhD required.