Wildberries reAnswer (Автоматические ответы на отзывы и вопросы WB)

An automated, AI-powered system designed for Wildberries marketplace sellers to manage, analyze, and reply to customer reviews and questions. Leveraging OpenAI GPT-4 and customizable rule engines, reAnswer automates customer support, maintains high ratings, and saves hours of manual work. It also provides messenger bot notifications (Telegram/MAX) and Yookassa payment integration.

---

🌟 Key Features

• AI-Powered Answers: Generates polite, brand-aligned, and context-aware responses to reviews and questions using OpenAI's GPT models.
• Custom Reply Rules: Setup granular rules based on product rating, keywords, specific products, and auto-reply templates.
• Real-time Notifications: Telegram and MAX bot integrations send instant updates and support interactive callback quick-actions.
• Admin Dashboard: Manage database entities, audit logs, and users easily through an integrated SQLAdmin interface.
• Monetization: Built-in Yookassa payment gateway support for user subscriptions and packages.
• Bilingual Interface: Seamless translation switcher supporting both Russian and English locales.

---

🛠 Tech Stack

Frontend

• Framework: Next.js 16 (App Router) & React 19
• Styling: Tailwind CSS v4, Framer Motion (for smooth micro-animations)
• State Management: Zustand
• Icons & FX: Lucide React, Canvas Confetti

Backend & Core Services

• API Framework: FastAPI (Python 3.11)
• Database & ORM: PostgreSQL 15, SQLAlchemy 2.0
• Database Admin: SQLAdmin 0.26 (WTForms-based admin panel)
• AI Engine: OpenAI Python Client (GPT-4 / GPT-5 reasoning models support)
• HTTP Clients: httpx, urllib3
• Authentication: JWT (JSON Web Tokens), bcrypt hashing

Reverse Proxy & Deployment

• Web Server: Nginx (handling reverse proxy routing to frontend and backend)
• Containerization: Docker & Docker Compose

---

📁 Directory Structure

├── src/                      # Next.js frontend source code
│   ├── app/                  # Frontend pages & layouts (landing, dashboard, settings, etc.)
│   ├── components/           # Reusable UI components
│   ├── hooks/                # Custom React hooks
│   ├── lib/                  # Utility libraries & API client helpers
│   ├── locales/              # Translation dictionaries (ru.ts, en.ts)
│   └── store/                # Zustand global state stores
├── backend/                  # FastAPI backend source code
│   ├── processor/            # Core business logic processors (controller, GPT integration)
│   │   ├── controller.py     # Main loops to fetch and process reviews & questions
│   │   └── gpt.py            # Async wrapper for OpenAI chat completions
│   ├── routers/              # API router endpoints (auth, rules, reviews, settings, etc.)
│   ├── services/             # Core service integrations (Yookassa, notifications, products)
│   ├── database.py           # DB connection setup
│   ├── models.py             # SQLAlchemy DB schemas
│   ├── controller_runner.py  # Script to launch the background processing task
│   └── bot.py                # Telegram & MAX bot runner and webhooks handler
├── nginx/                    # Reverse proxy configurations
│   └── default.conf          # Nginx routing rules mapping /api and /admin to backend
├── Dockerfile                # Frontend Docker build
├── docker-compose.yml        # Multi-container service orchestrator
└── backend/Dockerfile        # Backend Docker build

---

⚙️ Configuration & Environment Variables

Create a .env file in the root directory. Example configuration:

# Domain & API URL
DOMAIN=reanswer.ru
NEXT_PUBLIC_API_URL=https://reanswer.ru/api

# OpenAI API Key
OPENAI_API_KEY=your-openai-api-key

# Database Settings
DB_USER=user
DB_PASSWORD=your-secure-password
DB_NAME=autoreviews
DB_URL=postgresql://user:your-secure-password@db:5432/autoreviews

# Telegram & Messenger Bots
TG_BOT_TOKEN=your-telegram-bot-token
TG_BOT_NAME=your_telegram_bot_username
MAX_BOT_TOKEN=your-max-bot-token
MAX_BOT_NAME=your_max_bot_username
BOT_WEBHOOK_BASE_URL=https://reanswer.ru
BOT_WEBHOOK_SECRET=your-webhook-secret-token

# Admin panel
SQLADMIN_USERNAME=admin
SQLADMIN_PASSWORD=secure-admin-password
SQLADMIN_SECRET_KEY=some-random-secret-key

# YooKassa (Payments)
YOOKASSA_SHOP_ID=your-shop-id
YOOKASSA_SECRET_KEY=your-yookassa-secret-key

---

🚀 Getting Started

Method 1: Running with Docker Compose (Recommended)

To launch the entire application stack (Frontend, Backend, Database, Controller, Bots, and Nginx proxy) in production or local simulation mode:

1. Clone the repository and navigate into it.
2. Fill in the .env file.
3. Run the docker containers:

   docker compose up --build

4. Access the application in your browser:
   • Web App: http://localhost:8082 (or the configured DOMAIN)
   • Admin Panel: http://localhost:8082/admin (using SQLADMIN_USERNAME and SQLADMIN_PASSWORD)

---

Method 2: Local Development Setup

If you prefer to run services individually for debugging:

1. Database Setup

Ensure PostgreSQL is running and create a database matching your .env configuration.

2. Backend & Processes

Create a virtual environment, install requirements, and run the FastAPI server:

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

# Start API server
uvicorn backend.main:app --host 127.0.0.1 --port 8000 --reload

In separate terminals, run the controller and the messenger bot client:

# Start background worker/controller (fetches and replies to reviews)
python -m backend.controller_runner

# Start Telegram & MAX bot runner
python backend/bot.py

3. Frontend Setup

Install npm dependencies and launch the Next.js development server:

# From the root directory
npm install
npm run dev

Open http://localhost:3000 to access the frontend development build.

---

🤖 Bot Setup & Webhooks

When the FastAPI server starts, it registers webhooks automatically with Telegram/MAX platforms if BOT_WEBHOOK_BASE_URL and BOT_WEBHOOK_SECRET are configured.

1. Ensure your server is accessible via public HTTPS (e.g. through Nginx with SSL, Cloudflare, or Ngrok during local development).
2. Bot callback queries are forwarded to the /api/bot/webhook/{bot_type}/{secret} endpoint.
3. Interaction events will trigger immediate system replies or notify target admins.

---

📝 License

This project is proprietary. All rights reserved.