Stealth browser scraping — faster imports, no Bright Data bill

Under the hood, our scrape pipeline tries three tiers in order:

1. Crawl4AI (fast, for un-blocked sites)
2. Web Unlocker (our proxied fetch for Cloudflare-protected sites)
3. Stealth browser (full headless browser for sites that need JS rendering)

We replaced the old third-tier (Bright Data's managed Browser API) with an in-house stealth browser endpoint. It's faster, has no per-minute pricing, and gives us better control over fingerprinting.

What you'll notice: imports complete in seconds, not tens of seconds, on most platforms. And we can add new sites without blowing through a SaaS quota.