Acre AI: making land comps legible.

Land investors judge a deal by comparing it to recent nearby sales. The workflow was scattered across listing sites, county records and manual spreadsheets. Slow, error prone, and impossible to do quickly on a phone while standing on a lot.

Acre needed to pull that whole judgment into one product: find the property, surface the comparable sales, and show whether the numbers work, without burying an investor in data they have to interpret themselves.

The core call I made: the interface should answer a question, not present a table. Every screen leads with the thing an investor actually decides on, and pushes the raw data one layer down for when they want to verify it.

I designed the comping workflow so the comparison reads top to bottom like an argument. Subject property first, comps ranked by relevance, then the margin. The desktop version gives analysts room to dig; the mobile version compresses the same logic to a thumb-reachable summary.

This is the funding calculator's logic, rebuilt as a working component so you can feel the interaction rather than look at a flat screen. Drag the inputs. The margin and the go/no-go call update the way they do in the product.

The production calculator handles the full funding picture: purchase, costs, funding source and the resulting position. I kept the visual weight on the outcome so an investor reads the verdict first and the inputs second.

Investors evaluate lots on site. The mobile comping flow keeps the desktop's logic but strips it to what you can act on with one thumb: is this property worth pursuing, yes or no.