Skip to main content

Project Information Fields

A Project Information Field is a custom metadata field attached to a project — something like "Building Class", "Lot Number", "DA Reference", or "Purpose". Fields give you a structured place to record information that:

  • The approver needs to reference during review.
  • Should flow into generated project documents via merge fields.
  • Can be extracted from uploaded documents automatically by AI.

Fields are configured under Project Creation Setup → Project Information Fields (/project-creation-setup/project-information-fields) and shipped into projects via Project Types.

Two flavours: project-level and per-document​

Every field is one of two flavours, determined by whether it is linked to a document set.

Project-level fields (no document set)​

A field that is not linked to a document set has one value per project. Examples: "Building Class", "Site Address", "DA Reference". These are the default — leave the "Document Set" binding empty and the field behaves at the project level.

Per-document fields (linked to a document set)​

A field that is linked to a document set has one value per document in that set. Examples: "Purpose" linked to "Relied Upon Documents" — each document in that set gets its own Purpose value.

Per-document fields are what let you build document-template tables where each row corresponds to a document with its own column values. They're also how you record classification metadata alongside each received document without inventing ad-hoc note fields.

You decide the flavour when you define the field in the library. If you change the binding later, only new projects pick up the change.

AI extraction via a prompt​

Each library field has an optional AI Prompt (called the extraction prompt) that tells ApprovIQ how to propose a value for that field from the project's documents. For example, a DA Reference field might carry:

"Find the Development Application reference number. It is usually formatted like DA-2025-123 and appears in the top right of the first page of the consent document."

Extraction runs in two situations:

  • At project creation, ApprovIQ reads the documents the applicant has already uploaded in the background and proposes values for every field that has an extraction prompt. See Creating a Project.
  • On demand, when an approver or co-approver clicks Extract on a field.

Extraction runs for both project-level and per-document fields. For a per-document field, ApprovIQ fans out across every document in the linked set in parallel, proposing a value per document. Nothing is filled in silently — every proposal is shown for review and acceptance.

Writing a good extraction prompt is the same skill as writing a good document check: be specific about where the value lives and how to format it.

Editing field values manually​

You can always set field values manually instead of — or in addition to — AI extraction. Project-level values are edited on the project's information page. Per-document values are edited on the project's document-set page alongside each document. Any value you set manually overrides earlier AI proposals for that field and document.

Who can edit fields and run extraction​

Editing field values and triggering extraction on a project are available only to approvers and co-approvers of that project. Applicants and collaborators can see values surfaced in generated documents, but cannot change them or run extraction. The library itself is per-user: each approver maintains their own library of field definitions and prompts.

Using field values in generated documents​

Once values exist on a project, they become available as placeholders for document templates:

  • Project-level fields are available as project-wide merge fields under four names: «ProjectInformation_<name>», «ProjectInformation_<nameSpaceless>», «<name>», and «<nameSpaceless>». All four produce the same value — use whichever reads best in the template. The prefixed spaceless form is the safest default.
  • Per-document fields appear as per-row merge fields inside the document set's TableStart / TableEnd region in a table — so a body row wrapped with «TableStart:DocumentSet_ReliedUponDocuments» and «TableEnd:DocumentSet_ReliedUponDocuments» can reference «Purpose» directly, and the row is repeated once per document in the set with each document's own Purpose value.

In addition, two built-ins are always available: «ProjectName» and «ProjectAddress».

The workflow Generating a Project Document from Relied-Upon Documents walks through the full setup.