PetTalk

"Talk to Your Pet – An AI Pet Avatar That Responds with Personality With your Question and Answer !"

Comment

🐾 Inspiration

The idea for PetTalk was born from observing how young children bond with their pets — talking to them, asking them questions, and often imagining them responding back. We wanted to recreate that joyful experience in a digital way to support early learning. By combining AI with personalization, we aimed to create a tool where a child’s favorite pet becomes their study buddy and fun knowledge companion.

🧠 What it does

PetTalk is an interactive educational web app where children can:

  • Upload a picture of a pet (real or imagined).
  • Give it a name and select a personality (e.g., funny, serious, genius).
  • Ask any question — general knowledge, learning topics, or fun facts.
  • Get a text answer styled to the chosen personality, along with a unique voice/audio response matching that personality's tone.

For example:

  • A funny cat might answer in a playful voice with jokes.
  • A genius dog might give a detailed, calm answer.
  • A cheerful bird might use an excited tone with sound effects.

This combination of visual, textual, and audio feedback makes learning feel more magical and personal — especially for young users.

🛠️ How we built it

  • Frontend: React.js with Tailwind CSS and SCSS for styling.
  • Photo & Personalization: Users upload pet images, name them, and choose from preset personalities.
  • Database: Used Supabase to store user information and pet preferences.
  • AI Response Engine: Integrated Microsoft Azure OpenAI GPT-4 API to generate smart, friendly answers.
  • Voice Output: Used Microsoft Azure Speech API to generate text-to-speech audio based on personality.
  • No Animations: We deliberately kept it simple and focused on text + voice interactions without any animated elements.

🚧 Challenges we ran into

  • Matching voice tone and speech style to each personality in a consistent way.
  • Ensuring all content (text and audio) remains safe, clear, and appropriate for children.
  • Designing an interface that is intuitive for kids but also engaging.
  • Handling different accents and pronunciations in voice generation.

🏆 Accomplishments that we're proud of

  • Created a seamless pipeline from image upload to voice output.
  • Successfully mapped multiple personalities to different text + audio styles.
  • Used Supabase to manage user sessions and personalized pet settings.
  • Delivered both accessibility (audio for non-readers) and interactivity in a simple app.

📚 What we learned

  • How to humanize AI with personality-driven responses.
  • How to work with TTS (Text-to-Speech) tools to provide custom voice outputs.
  • How to balance creativity and safety when designing for young users.
  • The importance of multi-modal learning (text + voice) in engaging children.
  • How to use Supabase as a simple yet powerful backend service for user data.

🚀 What's next for PetTalk

  1. Add more voices and personalities for greater variety.
  2. Introduce a voice input option — so children can talk to their pet instead of typing.
  3. Implement a reward system (stickers, stars) to motivate learning.
  4. Build parental dashboards to track topics and learning progress.
  5. Add multilingual support so children can interact in their native language.

Built With

Share this project:

Updates