Apps Directory

Apps Directory

This directory contains user-facing applications for the CC-TPO project.

Applications

### 🎨 liquid-chat-ui
Next.js-based chat interface with liquid visual effects and intelligent conversation flow.

Tech Stack:
- Next.js (App Router)
- TypeScript
- Tailwind CSS
- Prisma + SQLite
- OpenAI GPT-4o

Features:
- Liquid ring topology visualization
- AI-powered chat with IRCP context
- Real-time conversation updates
- Prisma-based message storage

Development:

cd liquid-chat-ui
npm install
npm run dev

---

### 🔍 ircp-search-app
IRCP-powered semantic search application with ring topology context management.

Tech Stack:
- Next.js (App Router)
- TypeScript
- OpenAI GPT-4o-mini
- Python backend integration

Features:
- Semantic search with IRCP embeddings
- Ring context pruning for efficient token usage
- Real-time AI chat with context awareness
- Emotion analysis

Development:

cd ircp-search-app
npm install
npm run dev

---

### 🚀 liquid-chat-backend
FastAPI backend for IRCP processing and semantic search.

Tech Stack:
- FastAPI (Python)
- PyTorch
- Custom SentenceTransformer model
- SQLite

Features:
- Custom IRCP model for embeddings (384-dim)
- DLM coordinate calculation
- WebSocket support for real-time updates
- Semantic search API

Development:

cd liquid-chat-backend
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt
python main.py

---

Architecture

All applications share access to:
- Packages (`/packages/`): Core IRCP, TPO, RCP libraries
- Services (`/services/`): Shared backend services (search-api)
- Data (`/data/`): Databases and datasets

Environment Variables

Each application requires its own `.env` file:

liquid-chat-ui/.env:

OPENAI_API_KEY=your_key_here
NEXT_PUBLIC_OPENAI_MODEL=gpt-4o
DATABASE_URL="file:../data/databases/liquid-chat.db"

ircp-search-app/.env.local:

OPENAI_API_KEY=your_key_here

liquid-chat-backend/.env:

# Model paths configured in code