AI Telegram Bot for Users with RAG and Voice Messages

Developed an AI-powered Telegram bot that combines text communication, document-based Q&A, and voice messages within a single interface.

Project features:
— contextual chat with short-term conversation memory
— question answering over uploaded documents (PDF, DOCX, TXT) using RAG
— voice message processing: speech recognition, query analysis, and text/voice responses
— user and chat data isolation
— Docker-ready deployment setup
— support for both polling and webhook modes
— health checks, basic testing, and CI

Technologies:
Python, aiogram, LiteLLM, Google Gemini, Qdrant, Redis, FastAPI, Docker, Whisper, Edge-TTS.

The project was built as a ready-to-use foundation for an AI service in Telegram that can be scaled for customer support, FAQ automation, internal knowledge bases, or document workflows.

https://github.com/makquella/telegram-bot-ai-support

Tags
#TelegramBot #Python #AI #RAG #Gemini #Redis #Qdrant #FastAPI #Docker #Chatbot