Architecture of an Anti-Bot Protection
summary
Describes how modern anti-bot systems detect scrapers using IP/proxy checks, TCP/TLS/HTTP2 fingerprinting (client_hello/JA3-style), browser-side JavaScript and canvas/GPU fingerprints, behavior tracking and ML classification. Shows a proxy aggregator called Scrapoxy that can spin thousands of AWS proxies and reports sustained usage, and advises layered evasion using headless or real browsers (Puppeteer/Selenium/Playwright), mixed proxies, and monitoring 100–200 signals (TLS, headers, WebRTC leaks, canvas) to reduce false positives.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 1:48Project origin Scrapoxy
- 4:15Turing test problems
- 6:21AntiBot layers analysis
- 17:58False positives and limits
- 20:45Bypassing methodology
- 24:35Web scraping architecture
- 33:15Site-specific adaptation
- 35:01Obfuscation challenges
- 43:31Conclusion and demo
related talks
34mArchitecture of an Anti-Bot Protection
43mArchitecture of an Anti-Bot Protection
47mArchitecture of an Anti-Bot Protection
42mCracking the Code: Decoding Anti-Bot Systems!
23mBot or human? Detecting malicious bots with machine...
46mMastering Web Scraping: Unleash Your Data Extraction Wizardry!
40mRock 'Em Sock 'Em Robots: Bot Swatting Like the Pros
31mStrategies for Identifying and Mitigating Bot Threats