Ana Sayfa / Kod ve Scriptler / AI Contact Center Telegram Bot with Knowledge Base

AI Contact Center Telegram Bot with Knowledge Base

Bu Telegram botu, destek ekiplerinin belgelerinde yapay zekâ yanıtları bulmasına ve SSS'leri, şablonları ile programları yönetmesine yardımcı olur.

  • Yayın tarihi17 Kasım 2026
  • Nerede çalışırPython 3.12
  • Paket içeriği.py, .json, .md, .yml/.yaml, Dockerfile
  • Son güncelleme17 Kasım 2026
  • Canlı demoVar
AI Contact Center Telegram Bot with Knowledge Base

Overview

AI Contact Center Telegram Bot helps support teams find answers faster. Operators ask a question in Telegram and get an answer from YOUR instructions in seconds — with a button to open the original document. Team leads publish news and documents right in the chat; the AI indexes them instantly and notifies the team.

KEY FEATURES • AI knowledge base (RAG, Google Gemini): upload texts, PDF, DOCX, XLSX, PPTX, CSV, HTML and photos directly in Telegram. Answers come from your documents, with 📎 buttons to open the originals; an optional general-knowledge fallback is clearly marked and can be turned off. • FAQ with instant answers + log of unanswered questions (the best source for new FAQ entries). • Message templates with fields to fill in, per interface language, with preview. • Work schedule: on/off duty with break time tracking. • Announcements with "Acknowledged" tracking. • Access management: join requests, approval, roles (User / Administrator / Moderator), blocking and restoring, several departments (lines). • 📖 Read-only document library for operators. • Request distribution: split an Excel/CSV list of requests evenly among operators; result as Excel. • Excel exports of all data. • 3 interface languages: English, Ukrainian, Russian (each user chooses; add a language = one JSON file). • Gemini models are selected automatically and replaced without code changes when Google retires a model. • Activity log for easy support (never logs the text users type).

TECHNOLOGY Python 3.12, python-telegram-bot, Google Gemini API, Google BigQuery, FAISS. Docker, docker-compose and a Render blueprint included. 85+ automated tests.

REQUIREMENTS • Telegram bot token (free, @BotFather) • Google Cloud project with BigQuery (usually within the free tier) and a Gemini API key (pay per use) • Hosting for a Python/Docker worker: Render (~$7/month), any VPS or your own server

WHAT YOU GET • Full, readable and editable source code (no encryption, no license checks) • Step-by-step installation guide, user guide, security & data guide — in English, Ukrainian and Russian • Configuration templates (.env.example, Dockerfile, docker-compose, render.yaml) • Automated test suite

SUPPORT 3 months of support via Codester messages: questions about the product and help with product errors. Installation and configuration are done by the buyer following the documentation (typically 30–45 minutes).

Features

  • AI knowledge base (RAG, Google Gemini) — upload texts, PDF, DOCX, XLSX, PPTX, CSV, HTML and photos right in Telegram; answers come from your documents, with 📎 buttons to open the original (an optional general-knowledge fallback is clearly marked and can be turned off)
  • Instant indexing — new documents are searchable in seconds; the team gets an AI summary of every new document
  • FAQ with instant answers and a log of unanswered questions
  • Message templates with fields to fill in, per interface language, with preview before saving
  • Work schedule — on/off duty with break time tracking
  • Announcements with "Acknowledged" tracking
  • Access management — join requests, approval, roles (User / Administrator / Moderator), blocking and restoring, several departments (lines)
  • Read-only document library for operators
  • Request distribution — split an Excel/CSV list evenly among operators, result as Excel
  • Excel exports of all data (schedule, unanswered questions, acknowledgements, FAQ, templates, users)
  • 3 languages — English, Ukrainian, Russian; every user chooses; a new language = one JSON file
  • Self-updating AI models — the newest Gemini model is selected automatically and replaced without code changes when Google retires a model
  • Google BigQuery storage — tables are created automatically, embeddings are stored (no re-indexing after restart)
  • Activity log for easy support (never logs the text users type)
  • Docker, docker-compose and Render blueprint included; 85+ automated tests
  • Clean, readable, fully editable Python code — no encryption, no license checks

Requirements

  • Python 3.12+ or Docker
  • Telegram bot token (free, from @BotFather)
  • Google Cloud project with billing enabled:
    • BigQuery API + service account (roles: BigQuery Job User, BigQuery Data Editor) — usually within the free tier
    • Gemini API key from Google AI Studio — pay per use
  • Hosting that can run a Python/Docker worker 24/7: Render (~$7/month), any VPS or your own server
  • No public domain or SSL needed (long polling); webhook mode is optional