CS2 Cheat Detector — AI-Powered Telegram Bot
Project Description
Telegram bot for automatic analysis of CS2 game demo files using artificial intelligence. The system conducts deep behavioral analysis of players and detects cheating with an accuracy of up to 89%.
Key Features
Telegram Bot:
Automatic upload and processing of demo files (.dem)
Limit system: 3 free analyses + referral program
AI analysis of player behavior using Claude API
Interactive menu with inline buttons for navigation
Detailed reports with suspicion categories
Web Admin Panel:
User management and their analysis balance
Referral system with statistics tracking
Monitoring of uploads and service usage
Configuration of limits and bot parameters
Analysis System:
Parsing game events (kills, deaths, headshots)
Behavioral analysis: aim, reactions, gaming patterns
Evaluation of K/D, headshot percentage, game stability
Detection of suspicious moments with timestamps
Final verdict with probability percentage of cheating
Technology Stack
Backend:
Python - main development language
Flask - web server and API
pyTelegramBotAPI - integration with Telegram
Celery - asynchronous task processing
PostgreSQL - storage of user data and statistics
AI & Processing:
Claude API - neural network analysis of demo files
CS2 demo file parsers
Behavioral detection algorithms
Infrastructure:
Railway.app - cloud hosting
Gunicorn - WSGI server with concurrency settings
Nginx - proxying and load balancing
Docker - containerization of services
Implemented Functionality
For Users:
Upload demos via drag-and-drop interface
Real-time tracking of analysis progress
Receiving detailed reports
Help system with instructions
Referral system for earning bonus checks
For Administrators:
Flask admin panel with authorization
User database management
Usage statistics and analytics
Configuration of limits and rates
Security and Performance
Processing up to 40+ simultaneous users
API protection from unauthorized access
Optimized file uploads (up to 300MB)
Shared volume between containers for file exchange
Timeouts and graceful shutdown for stability
Results and Metrics
Analysis accuracy: ~89% (comparable to professional anti-cheats)
Processing speed: 2-3 minutes per demo file
Scalability: support for dozens of simultaneous analyses
User experience: intuitive interface with step-by-step feedback
Unique Features
AI integration for analysis instead of signature methods
Telegram as a platform - access without app installation
Monetization through limit and referral systems
Full cycle from upload to detailed report in minutes
Technical Details
The project demonstrates full development skills:
Microservices architecture (bot + web + worker)
Working with file systems and streaming
Integration of external AI APIs
Database design and migrations
Deployment and DevOps practices
UX design for messengers
Technologies: Python - Flask - Telegram Bot API - Claude AI - PostgreSQL - Celery - Docker - Railway - Gunicorn
Telegram bot for automatic analysis of CS2 game demo files using artificial intelligence. The system conducts deep behavioral analysis of players and detects cheating with an accuracy of up to 89%.
Key Features
Telegram Bot:
Automatic upload and processing of demo files (.dem)
Limit system: 3 free analyses + referral program
AI analysis of player behavior using Claude API
Interactive menu with inline buttons for navigation
Detailed reports with suspicion categories
Web Admin Panel:
User management and their analysis balance
Referral system with statistics tracking
Monitoring of uploads and service usage
Configuration of limits and bot parameters
Analysis System:
Parsing game events (kills, deaths, headshots)
Behavioral analysis: aim, reactions, gaming patterns
Evaluation of K/D, headshot percentage, game stability
Detection of suspicious moments with timestamps
Final verdict with probability percentage of cheating
Technology Stack
Backend:
Python - main development language
Flask - web server and API
pyTelegramBotAPI - integration with Telegram
Celery - asynchronous task processing
PostgreSQL - storage of user data and statistics
AI & Processing:
Claude API - neural network analysis of demo files
CS2 demo file parsers
Behavioral detection algorithms
Infrastructure:
Railway.app - cloud hosting
Gunicorn - WSGI server with concurrency settings
Nginx - proxying and load balancing
Docker - containerization of services
Implemented Functionality
For Users:
Upload demos via drag-and-drop interface
Real-time tracking of analysis progress
Receiving detailed reports
Help system with instructions
Referral system for earning bonus checks
For Administrators:
Flask admin panel with authorization
User database management
Usage statistics and analytics
Configuration of limits and rates
Security and Performance
Processing up to 40+ simultaneous users
API protection from unauthorized access
Optimized file uploads (up to 300MB)
Shared volume between containers for file exchange
Timeouts and graceful shutdown for stability
Results and Metrics
Analysis accuracy: ~89% (comparable to professional anti-cheats)
Processing speed: 2-3 minutes per demo file
Scalability: support for dozens of simultaneous analyses
User experience: intuitive interface with step-by-step feedback
Unique Features
AI integration for analysis instead of signature methods
Telegram as a platform - access without app installation
Monetization through limit and referral systems
Full cycle from upload to detailed report in minutes
Technical Details
The project demonstrates full development skills:
Microservices architecture (bot + web + worker)
Working with file systems and streaming
Integration of external AI APIs
Database design and migrations
Deployment and DevOps practices
UX design for messengers
Technologies: Python - Flask - Telegram Bot API - Claude AI - PostgreSQL - Celery - Docker - Railway - Gunicorn
Wrocław 
Available for hire