Content Understanding
Identifies contract data and checkbox values accurately. Every section is analysed — from special conditions to planning certificates — and a generated page map lets you navigate effortlessly.
Open a property contract PDF and the work begins before you do. Names, dates, checkbox states, special conditions, planning certificates, dealings — the data has to be off the page and into your review fields before you can think clearly about any of it. Curia's property contract data extraction handles that part silently, then hands you a contract you can navigate by section instead of by scroll bar.
The problem with reading a contract cold
A property contract is 60 to 200 pages of mixed material — the front of the contract, the standard terms, special conditions, vendor's statement, planning certificates, title searches, dealings, plans, leases. The structure is consistent enough that a conveyancer knows where to look, and chaotic enough that finding the right page still costs minutes per file. Pulling values out by hand — whether the deposit is 10% or less, which checkboxes are ticked on the front page, what each dealing actually is — is the slow tax on every review.
How Curia reads the contract
Every page is classified. Curia's automated contract analysis identifies what each page is — front of contract, special conditions, planning certificate, title search, dealing, plan, Section 32 statement — and groups them into the document's real sections. Multiple instances of the same document type (three title searches, several dealings) each get their own indexed section, so they don't blur together.
Form fields and checkboxes are parsed accurately. Curia reads the front-of-contract data and the checkbox states the way they're written on the page — deposit amount, settlement period, GST treatment, inclusions, the lot — and writes them straight into the review fields you're about to step through.
Special conditions are extracted in document order. Curia walks the special conditions clause by clause, captures the categories that matter, and keeps a reference back to the page and clause number for every entry. When you check one, the source is one click away.
A page map is generated. Once classification is done, Curia builds a navigable map of the contract — labelled by section, indexed by instance, with dealing numbers and categories where they apply. You jump to any section in a click instead of scrolling.
Built for Australian property contracts — NSW and VIC
Conveyancing document parsing isn't generic OCR; the structure of an Australian contract is specific. Curia is built around the NSW standard form contract — its front page, its prescribed documents, the way LRS dealings and plans are laid out — and around Victorian custom contracts and Section 32 statements , where the contract shape varies firm to firm and the vendor's statement is a separate document with its own structure. The extraction adapts to each.
Where it sits in the review
Content understanding is the data-in layer. It runs the moment you upload, before you start your walkthrough. By the time you open the file, the review fields are populated, the special conditions are listed in your firm's order, and the page map is ready. You spend your time on judgement, not on data entry — and the Smart Prefill and Workflows features take it from there.