What I Built
Bringing a new dog home is exciting... and also confusing.
Why are they crying?
Why are they hiding?
Why won't they eat?
Why are they biting my feet? 😭
I built PawPal, a playful AI companion for new dog owners that helps turn those confusing moments into simple, reassuring guidance.
I especially wanted PawPal to feel useful for people welcoming a rescued dog into their home, where a new environment can be overwhelming for both the dog and the person trying to understand them.
You describe what's happening, and PawPal explains what might be going on, suggests a few things you can try, and gives you a playful "What your dog really means" translation.
And then... you can actually hear your dog say it. 🐶🔊
PawPal isn't meant to diagnose a dog or replace a veterinarian. It's a friendly first step for a confused human trying to understand their new companion.
Demo
🐾 Try PawPal:
https://pawpalweb.netlify.app/
Try asking:
"My rescued puppy keeps hiding under the bed. What should I do?"
Then listen to "What your dog really means." 🐶
Code
The project is open source:
https://github.com/ishika-thakur7802/pawpal
How I Built It
PawPal is a small React/Vite application backed by Netlify serverless functions.
The basic flow is:
User → React UI → Netlify Function → Google Gemini → PawPal response
For the voice experience:
PawPal response → ElevenLabs → audio → user
I used Google Gemini to generate the personalized responses. The prompt gives PawPal a specific personality and structure so that responses stay concise, friendly, practical, and appropriate for someone who may be a first-time dog owner.
I then used ElevenLabs to turn the playful "What your dog really means" part into speech.
I kept the project intentionally small. Rather than building a huge pet-management platform, I focused on one interaction:
"I don't understand what my dog is doing. Help."
The rescued-dog prompts were also designed to require no additional backend complexity — they simply give new adopters useful starting questions.
Prize Categories
🌟 Best Use of Google AI
PawPal uses Google Gemini as the core reasoning layer.
Gemini takes the owner's description of what's happening and generates a personalized response with:
what might be happening
practical things the owner can try
when professional help may be appropriate
a playful interpretation from the dog's perspective
The AI isn't just decorating the interface — the main purpose of PawPal depends on it.
🌟 Best Use of ElevenLabs
ElevenLabs powers PawPal's voice experience.
After Gemini generates the dog's playful "translation," ElevenLabs turns it into audio so the user can hear what their dog is supposedly trying to say. 🐶
That was important to me because the interaction should feel fun rather than like filling out another AI form.
Why I Built It
I wanted to make something that could be genuinely useful during those first confusing days of having a new dog — especially when that dog has come from a rescue situation and both sides are still learning to trust each other.
Sometimes you don't need another complicated pet platform.
You just need someone to say:
"Okay. Your dog is probably trying to tell you something. Let's figure it out." 🐾
Built for the DEV Weekend Challenge: Dog Days Edition.
Top comments (0)