A property-tax engine that refuses to file a losing appeal
The problem
Independent landlords have no affordable way to tell whether their assessment is worth appealing. Enterprise data tools are priced for professionals, and contingency attorneys only engage above a value threshold most owners never reach.
The insight
The bill is not driven by the value you get to appeal
In Florida you appeal just (market) value, but the tax bill is computed on assessed value — just value reduced by the statutory cap. When assessed sits below just, an appeal that lowers just value but stays above assessed saves the owner exactly nothing. Naive comparison tools flag those properties anyway, and every one of those flags is a wasted filing fee and a lost customer. The engine was built around that split from the first commit, and the test suite pins it to hand-verified parcels. Allegheny County needed a second model entirely — assessments frozen at 2012 base-year values, where fairness is market × the Common Level Ratio rather than market itself.
What we built
- Ingestion for Florida DOR NAL/SDF roll files and the Miami-Dade ArcGIS parcel service, with folio normalisation across inconsistent county formats
- A cap-aware flag engine that separates three distinct verdicts: a legal case, savings this year, and strategic future-year value
- IAAO-standard ratio-study statistics to validate the comparable-sales model against the roll rather than trusting it
- Automated generation of the official DR-486 petition and DR-501T portability forms as filled AcroForms, with a comparables evidence packet
- A public instant-check flow that runs the full engine on public record data before asking for anything
Have something with a domain problem underneath it?
Those are the ones worth talking about. Describe it and you will get a straight answer about what it would take.
Start a conversation