A CMS where AI supports the writer, not replaces them

DoubleStack is a restaurant content site built around human-written stories. The food writing matters. The photography matters. AI generating the content was never the goal. The goal was building tools that let a writer publish faster and more consistently without losing what makes the writing theirs.

We built a custom CMS from scratch. When a writer wants to cover a new place, they search for it on Google Maps directly inside the editor. The CMS pulls in the address, proper name, and location data automatically. It classifies the place by city and fills in criteria. No manual data entry, no copy-pasting from Google.

Photography goes through an automated publishing pipeline. AI analyzes each image to find the focal point and crops accordingly. On the content side, AI determines cuisine types and food categories to build out the site's ontology. Every piece of content gets structured metadata without the writer having to think about taxonomy.

The built-in AI editor is where it gets interesting. Writers can request a critique at any point. The editor reviews the draft and gives feedback on writing voice, sentiment, hooks and takeaways, whether sensory details are fleshed out. It flags where the piece drifts from narrative into restaurant review territory. It doesn't write anything. It just tells you what needs work so the story stays a story.

The result is a content operation that scales without sacrificing voice. Every article reads like it was written by the same person with the same standards. Because it was. The AI just made sure of it.