Chat-Toner is an intelligent text conversion and analysis service designed to refine your communication. It helps you adjust the tone and style of your writing, ensures quality and consistency, and provides a knowledge base for company-specific communication protocols.
| Name | Role | |
|---|---|---|
| Yoon Jiwon | Project Lead | [email protected] |
| Ha Jimin | Frontend, UI/UX Research & Design | [email protected] |
| Jung Jieun | Infrastructure & Architecture Orchestration | [email protected] |
| Kim Jimin | RAG Construction, Langchain Agentic Flow & Learning System | [email protected] |
- Tone & Style Adjustment: Easily convert your text between different communication styles (e.g., formal, friendly, direct).
- Profile-Based Conversion: The system learns your preferences from surveys and feedback to provide personalized conversions.
- Comprehensive Scoring: Get scores for your text based on Grammar, Formality, Readability, and Protocol Compliance.
- Actionable Suggestions: Receive concrete suggestions for improving your text.
- RAG-Powered Justification: Understand why you received a certain score. The system uses its knowledge base to provide clear justifications for its analysis.
[Placeholder for a screenshot of the quality analysis results with justifications]
| Ingestion UI | RAG Q&A UI |
|---|---|
 |
 |
- Document Ingestion: Build a company-specific knowledge base by uploading documents (PDFs, etc.).
- Contextual Q&A: Ask questions and get answers based on the ingested documents, ensuring everyone follows the same guidelines.
- Onboarding Surveys: Quickly set up company-wide communication styles and protocols through a simple survey.
- Personalized Experience: User-specific profiles store preferences and feedback, making the tool more effective over time.
| Category | Technology |
|---|---|
| Frontend | React, TypeScript, Vite, Tailwind CSS, TanStack Query |
| API Gateway | NestJS, TypeScript |
| Backend (Core) | Python, FastAPI |
| AI & ML | LangChain, OpenAI (GPT-4), PGVector |
| Database | PostgreSQL (with PGVector extension) |
| DevOps | Docker, Docker Compose, GitHub Actions |
| Testing | Jest, Pytest |
| Code Quality | ESLint, Prettier, Ruff |
The Chat-Toner service is built on a microservices architecture, with a React frontend, a NestJS API gateway, and a Python backend for core AI/ML and business logic.
.
├── packages/
│ ├── client/ # React Frontend
│ ├── nestjs-gateway/ # NestJS API Gateway
│ └── python_backend/ # FastAPI Backend (Core Logic & RAG)
├── database/ # Database migration scripts
├── infra/ # Infrastructure configs (e.g., task definitions)
├── docker-compose.yml # Local development setup
└── README.md
- Node.js >= 18
- Python >= 3.10
- Docker & Docker Compose
-
Clone the repository:
git clone https://github.com/your-repo/2025-CHATTONER-Server.git cd 2025-CHATTONER-Server -
Set up environment variables: Create a
.env.localfile in the root directory and add the necessary environment variables. You can use.env.exampleas a template.cp .env.example .env.local
Key variables to set:
OPENAI_API_KEY: Your OpenAI API key.DATABASE_URL: The connection string for your PostgreSQL database.
-
Install dependencies for all packages: This project uses
npmworkspaces. Run the command from the root directory.npm install
This will install dependencies for the
client,nestjs-gateway, and set up the Python environment forpython_backend.
You can run all services together using Docker Compose.
docker-compose up --buildWe welcome contributions! Please follow these guidelines:
- Follow the commit conventions (e.g.,
feat:,fix:,docs:). - Ensure all tests and linting checks pass before submitting a PR.
- Update the
README.mdand any other relevant documentation if you make significant changes.
This project is licensed under the MIT License. See the LICENSE.md file for details.