AI Companion - Memory Extraction & Personality Engine

A simple web app that extracts user memories from chat messages and applies different personality tones to AI responses.

⚠️ Deployment Note

This project is hosted on Render’s free tier. If the backend has been idle for a while, Render puts the server to sleep. When it wakes up, the first request may take 20–30 seconds to load. After the initial spin-up, everything works normally.

A background uptime monitor keeps the service warm, but occasional delays are still expected on the free tier.

What it does

1. Memory Extraction - Give it 30 chat messages and it pulls out:

   • User preferences (likes/dislikes)
   • Emotional patterns
   • Important facts

2. Personality Engine - Takes a message and transforms the AI response based on tone:

   • Calm Mentor
   • Witty Friend
   • Therapist

Setup

Backend

cd backend
pip install -r requirements.txt
export OPENAI_API_KEY="your-api-key-here"
python app.py

Server runs on http://localhost:5000

Frontend

Just open frontend/index.html in your browser. Or use a simple server:

cd frontend
python -m http.server 8080

Then go to http://localhost:8080

API Endpoints

POST /extract_memory

Request:

{
  "messages": ["message 1", "message 2", "..."]
}

Response:

{
  "preferences": ["..."],
  "emotional_patterns": ["..."],
  "facts": ["..."]
}

POST /apply_personality

Request:

{
  "message": "I'm feeling stressed",
  "tone": "calm mentor"
}

Response:

{
  "original_message": "...",
  "before": "basic response",
  "after": "response with personality",
  "tone": "calm mentor"
}

Deployment

Render (Backend)

1. Push to GitHub
2. Create new Web Service on Render
3. Set root directory to backend
4. Add environment variable OPENAI_API_KEY
5. Start command: gunicorn app:app

Vercel/Netlify (Frontend)

1. Set root directory to frontend
2. Deploy as static site
3. Update API_URL in index.html to point to your backend

Tech Stack

• Backend: Flask (Python)
• Frontend: Plain HTML/CSS/JS
• LLM: OpenAI GPT-3.5

Notes

• You need an OpenAI API key to run this
• The app doesn't store any data
• Keep your API key secret!