The whole thing runs on PowerShell and AI. Feed it a domain list, it generates content, builds the pages and deploys to Vercel automatically. I just review the output.
Setting up the pipeline took time. Running it now takes hours not weeks.
This is what AI leverage actually looks like when you build the right systems around it. The bottleneck stopped being execution and became decision making.
Anyone else building automation pipelines like this?
The interesting shift isn't generating 580 landing pages with AI—it's that automation moved the bottleneck from execution to deciding which pages are actually worth creating. I'd keep validating whether the long-term advantage comes from publishing at scale or from consistently making better decisions about what deserves to exist in the first place.
Five hundred eighty pages makes selection the product. I'd publish them in cohorts of 20, then kill any cluster that gets impressions but no qualified click or conversion after a fixed window. Otherwise the pipeline optimizes page count while the review queue quietly becomes the new manual job.