Telegram Multi-Bot CRM + Analytics for Traffic Arbitration
Platform for managing Telegram bots with a CRM panel, automatic push notifications, and conversion analytics.
For traffic arbitrage teams across multiple geos (India, Latin America, post-Soviet countries excluding Ukraine).
What has been implemented:
— Multi-bot architecture: 4+ bots on one server through a single webhook server, each bot with its own configuration, prompts, and language.
— AI sports analyst: the bot generates previews of matches from top leagues (Premier League, La Liga, Serie A, Champions League) via Grok API + real-time search. Upon user request, it provides predictions for specific matches with analysis of lineups, team form, and statistics. Main monetization — conversion to deposits through the funnel, additional — selling predictions via Telegram Stars (star balance, daily free predictions).
— AI consultant for the Indian market: multilingual bot (English, Hindi, Bengali, Tamil) with auto language detection. LLM routing between Anthropic Claude, Groq, and xAI Grok depending on the bot.
— CRM panel: dashboard with real-time statistics, user pagination, complete chat history, filtering by bots and platforms, arbitrage funnel analytics (chats → registrations → deposits → conversions by days).
— Push system: automatic multi-level funnels (cron every 10 minutes), AI-generated text for each user, spam protection, cross-bot deduplication, auto-tagging of users who blocked the bot.
— Webhook integration with Keitaro: receiving postbacks of registrations and deposits, linking to users through affiliate tracking, idempotency in processing duplicates.
— Authorization with rate-limiting, managing bot configurations from the interface, logging critical errors, test data for debugging webhooks.
Stack:
Python · Flask · aiogram · SQLite (WAL) · nginx · systemd · asyncio · Anthropic API · Groq API · xAI Grok API · Telegram Bot API · Telegram Stars API · Keitaro · Git · Linux VPS
#Python #Flask #Telegram #TelegramBot #aiogram #CRM #arbitrage #traffic #AI #LLM #Anthropic #Grok #push #webhook #Keitaro #SQLite #nginx #Linux #asyncio
For traffic arbitrage teams across multiple geos (India, Latin America, post-Soviet countries excluding Ukraine).
What has been implemented:
— Multi-bot architecture: 4+ bots on one server through a single webhook server, each bot with its own configuration, prompts, and language.
— AI sports analyst: the bot generates previews of matches from top leagues (Premier League, La Liga, Serie A, Champions League) via Grok API + real-time search. Upon user request, it provides predictions for specific matches with analysis of lineups, team form, and statistics. Main monetization — conversion to deposits through the funnel, additional — selling predictions via Telegram Stars (star balance, daily free predictions).
— AI consultant for the Indian market: multilingual bot (English, Hindi, Bengali, Tamil) with auto language detection. LLM routing between Anthropic Claude, Groq, and xAI Grok depending on the bot.
— CRM panel: dashboard with real-time statistics, user pagination, complete chat history, filtering by bots and platforms, arbitrage funnel analytics (chats → registrations → deposits → conversions by days).
— Push system: automatic multi-level funnels (cron every 10 minutes), AI-generated text for each user, spam protection, cross-bot deduplication, auto-tagging of users who blocked the bot.
— Webhook integration with Keitaro: receiving postbacks of registrations and deposits, linking to users through affiliate tracking, idempotency in processing duplicates.
— Authorization with rate-limiting, managing bot configurations from the interface, logging critical errors, test data for debugging webhooks.
Stack:
Python · Flask · aiogram · SQLite (WAL) · nginx · systemd · asyncio · Anthropic API · Groq API · xAI Grok API · Telegram Bot API · Telegram Stars API · Keitaro · Git · Linux VPS
#Python #Flask #Telegram #TelegramBot #aiogram #CRM #arbitrage #traffic #AI #LLM #Anthropic #Grok #push #webhook #Keitaro #SQLite #nginx #Linux #asyncio
Kyiv 
A little busy