Intuned launches AI-generated Playwright automations with self-healing scrapers

Y Combinator S22 alum Intuned has launched a platform that turns natural-language prompts into production Playwright code for scrapers, crawlers, and RPA workflows. The Intuned Agent generates the automation, deploys it, and rewrites it when target sites change, while developers retain ownership of the underlying TypeScript or Python code with no vendor lock-in. Built-in features include anti-detection stealth, captcha solving, session and authentication management, proxy handling, scheduling, and auto-scaling across machines.

The platform spans four use cases: data extraction, large-scale crawling with Crawl4AI support, RPA for services lacking APIs, and hybrid AI automations that combine deterministic Playwright steps with agentic libraries like Anthropic Computer Use, OpenAI CUA, Stagehand, and Browser-use. A new Web Tasks API runs prompt-driven jobs that learn from prior executions — supplying a reuseKey caches the generated skill so subsequent runs are dramatically faster and cheaper, dropping from minutes and dollars to seconds and cents in the YC directory example.

For teams operating at scale, Intuned also offers a managed service where solution engineers paired with the agent build and maintain scrapers under flat per-source pricing. The pitch targets the perennial brittleness problem in web automation: getting code-level reliability and infrastructure without hand-writing or babysitting the selectors.