How Web Data Infrastructure Powers the Next Generation of AI — Patricija Žemaitytė, Oxylabs
AI models need real-time web data infrastructure, not just static training data
“training alone is no longer enough and to stay useful models needs to get access to fresh information, live search, real external data”
Oxylabs PM Patricija Žemaitytė argues that the AI industry is shifting away from relying solely on static training data toward requiring live, real-time web data pipelines. She frames web scraping infrastructure as a critical but unglamorous dependency for next-gen AI agents. The talk is primarily a vendor perspective with an anecdote about building a 5 petabyte/month video data pipeline in two weeks for an AI training customer.