This project provides a Telegram and Signal support bot designed for managing customer inquiries and providing automated assistance. It allows users to create tickets that are routed to a staff group, enabling efficient handling of support requests, and integrates LLMs for AI-powered responses.

How It Works

The bot functions by intercepting user messages in a support chat, creating a ticket, and forwarding it to a designated staff group. Staff members can respond by simply replying to the forwarded message. Key features include file forwarding, ticket database management, user restriction, anti-spam measures, and the ability to send tickets to multiple staff groups. The LLM integration allows for automatic responses to common questions or when staff are unavailable, enhancing support availability.

Quick Start & Requirements

• Install via Docker: docker-compose up -d
• Requires a config.yaml file, derived from config-sample.yaml.
• LLM integration requires API keys and configuration for OpenAI or compatible APIs.
• Official Wiki for detailed instructions: https://github.com/bostrot/telegram-support-bot/wiki

Highlighted Details

• Supports both Telegram and Signal platforms.
• Integrates with LLMs (OpenAI, OpenAI-compatible APIs) for automated responses.
• Features include file forwarding, ticket database, user restriction, and anti-spam.
• Offers private replies, user anonymization, and keyword-based auto-replies.

Maintenance & Community

The project is maintained by Eric Trenkel. Contributions, issues, and feature requests are welcome via the GitHub issues page. Support for a hosted solution is available at Botspace.

Licensing & Compatibility

• Licensed under GPL-3.0.
• The GPL-3.0 license may impose copyleft restrictions, requiring derivative works to also be open-sourced under the same license, which could affect commercial or closed-source integration.

Limitations & Caveats

The project has undergone breaking changes, notably the migration from telegraf to grammY in v4, requiring configuration updates. The LLM auto-reply feature is marked as beta.