AI Assistant Telegram| RAG Systems | Django+LangChain+OpenAI

AI Assistant (RAG Platform for User Support)
A fully Dockerized environment that can be quickly deployed for your business, the admin panel allows uploading documents with information for users to search. A knowledge base for managers, search across chats.
Additional configurations can be implemented to meet the specific needs of a business.

This is a complete AI system for supporting users and managers, built on RAG architecture. The project combines a Telegram bot, a manager's web cabinet, and AI search across internal documents. With chat statuses, internal notes for chats.

Technology Stack
Django, DRF, Django Channels, aiogram 3, LangChain, OpenAI, Qdrant, PostgreSQL, Redis, Celery, Docker Compose,

Capabilities
* AI assistant in Telegram with dialogue context support
* Responses based on internal documents (RAG + Qdrant)
* Live WebSocket chat between user and manager (takeover mode)
* FAQ system with quick responses
* Admin panel for uploading documents and indexing them
* Automatic creation of embeddings through OpenAI and LangChain
* Task queues via Celery (indexing, notifications)
* Chat status system (AI / waiting for manager / with manager)

Advantages
* Scalable AI architecture without microservices
* Fast responses through vector search (Qdrant)
* Ability for live manager intervention in chat
* Minimal costs due to optimized RAG
* Flexible document and FAQ system without releases

#Django #RAG #AI #OpenAI #LangChain #Qdrant #PostgreSQL #Redis #Celery #Docker #TelegramBot #WebSockets #DjangoChannels #FullStack #Backend #AIassistant #MachineLearning #HTMX #Python