Find the project. Find the task. Find the word on the page.

How it works

Global search in the dashboards

A static Search field sits in the top header on every dashboard. Search scope follows the dashboard the user is in (projects from Home or Projects, tasks from Tasks, files from Assets). The match is literal: "Summer Campaign" matches the full string, not the words separately. The 2025/R1 release added the ability to search within an active filter, so users can narrow further inside a filtered view rather than starting over. The dashboard search rules are covered in detail in the Dashboards and Project Organization card.

Advanced filters: standard mode

Standard mode builds filters through a condition editor. The user picks a metadata key from the dropdown of keys their permission exposes, picks an operator (=, contains, is empty, is not empty, and others depending on the key type), and enters a value. Multiple conditions are combined with All (AND) or Any (OR) logic. Each filter is named, described, and can be previewed for match count before saving.

A filter can be set up against a custom metadata key (campaign code, region, brand, regulatory class) so the filter answers a business question the platform's built-in filters cannot. Data fields used in the filter can be promoted to dashboard columns so the filtered view shows the right context.

Advanced filters: expert mode (AQL)

Expert mode opens the full Aproove Query Language. AQL queries are written directly with dropdown helpers for system attributes, operators, Tags, and metadata keys. The language supports:

• System attributes for projects (ProjectName, ProjectDeadline, ProjectStartDate, ProjectStatus, ProjectTag, ProjectTreePath, ProjectCreatorID, TasksLate, TasksAtRisk, ActiveStepName, NbProof, NoteContent, and many others).
• System attributes for tasks (IsWorkflowTask, IsAdhocTask, IsMention, IsExpired, IsDone, IsConflict, AssignedToUserId, AssigneeReportsTo, Decision, DateAssigned, ActiveStepDeadline, and others).
• Dynamic variables that resolve at runtime to the connected user: Me (the user contact object), MyId, MyEmail, MyLogin. Date variables include Now, CurrentDate, CurrentTime, with arithmetic support (NOW - 7, CURRENTDATE + 14).
• Custom metadata keys with type-appropriate operators (text keys use =, contains, like, is empty; date keys use <, >, =, !=, ranges; boolean keys use =).
• Sorting and comments. Order by ASC or DESC on one or more fields. Comments inside /* */ are ignored at execution; tag names appear automatically as inline comments when inserted from the Tags dropdown, so a saved query like ProjectTag = 4 /*Approved*/ stays readable a year later.

A query like ProjectCreatorID = MyId AND ProjectDeadline <= (NOW + 7) AND IsNotArchived order by ProjectDeadline ASC becomes the saved view "My projects due in the next week," available to the user as a Favorite, shareable to their team, and updatable as their workload changes.

Sharing advanced filters

Saved filters can be shared with individual contacts or contact groups at three permission tiers:

• Read only (default): the recipient can use the filter but cannot edit or share it further.
• Allow Sharing: the recipient can use the filter and share it with others, but cannot edit it.
• Admin: the recipient has the same rights as the filter creator (edit and share).

Shared filters surface in the recipient's Favorites and behave the same as their own filters.

Full-text content search inside proofs

In Proof Plan View, clicking the magnifying glass opens a search field that runs across the actual content of every proof in the project. The search covers:

• Page content across all proof versions. The count under each proof thumbnail is for the current version only; occurrences on previous versions are shown when the proof is opened but do not contribute to the count.
• File names.
• Notes on the proof (this can be toggled off in user settings if notes should be excluded).

Search is case-insensitive (Francisco and francisco return the same results). The match is word-by-word rather than full-string ("Annual cost" returns occurrences of "annual" and "cost" separately).

Clicking a proof opens it and shows each occurrence. Clicking any occurrence zooms to it and centers it in the viewport.

Apply a note to every occurrence. The Select All button on the search results applies a single note to every occurrence found in the project, in one action. A reviewer who spots a misspelling, an outdated product code, a deprecated brand reference, or a flagged regulatory term can annotate every instance across every proof in the project at once, rather than navigating to each occurrence individually. This is the capability most browser proofing tools do not have, and it is where Full-Text Content Search becomes operational rather than informational.

Benefits

• Three layers of search for three different questions. Dashboard search finds projects and tasks. Advanced Filters build saved queries on workflow state. Full-Text Content Search finds words on the page.
• Saved filters, shareable to teams. Custom filters are saved once and shared with three permission tiers, so a team's preferred views travel with the team.
• Standard mode for everyday users. Point-and-click condition builder uses dropdowns for metadata keys, operators, and values. No syntax required.
• Expert mode for power users. Full AQL with system attributes, dynamic user and date variables, and date arithmetic, written directly with dropdown helpers so power users get speed and precision.
• Filter data fields become dashboard columns. Filter fields can be promoted to columns so the filtered view shows the right context, not just the right rows.
• Custom metadata in queries. Filters and AQL can reference any metadata key the user has permission to see, so business-specific dimensions (campaign code, brand, region, regulatory class) become first-class search criteria.
• Dynamic personal queries. Variables like Me, MyId, and MyEmail resolve at runtime, so a single filter can be "my projects" or "tasks assigned to people who report to me" without rebuilding the query for each person.
• Date arithmetic in filters. Queries like Deadline <= (NOW + 7) give the filter the right answer every time it runs, rather than the right answer the day it was built.
• Find content inside the proof, not just metadata about it. Full-Text Content Search reads the rendered text of every proof, so a reviewer can find every occurrence of a specific word, code, or phrase across the project without opening each proof individually.
• Batch annotation from search. Apply one note to every occurrence of the search term in the project, in one action. Misspellings, outdated codes, deprecated terms, and flagged language get noted everywhere at once.

Who it's for

• Project managers and operations leaders building saved filters for portfolio views (late projects, my projects, region-scoped projects, projects with conflict tasks).
• System administrators and power users writing AQL queries against system attributes and custom metadata for surgical filtering and reporting.
• QA and compliance reviewers using Full-Text Content Search to verify that flagged terms, deprecated codes, regulatory disclosures, or brand-restricted words appear (or do not) in every proof of a project.
• Editorial and copy review teams finding every instance of a misspelling, an outdated product name, or a term that needs replacement, and annotating them all at once.
• Brand and regulatory teams running consistency checks across many variants of the same campaign.

Under the hood

Advanced Filters are split into two types (project filters and task filters) and stored against the user who created them. Sharing permissions are stored against each shared assignment. AQL queries are parsed at execution time; dynamic variables resolve from the connected user's session; date arithmetic resolves from the server time.

Full-Text Content Search reads the extracted text content from Aproove's file processing pipeline (the same pipeline that powers AI Agent grounding, Page-Level Smart Review, and proof rendering). Searched objects include the rendered page text across all proof versions, the file name, and the notes attached to the proof. If page content search returns no results when text content is expected, the processing agent setting on the backend may need configuration.

AQL queries can be invoked from outside Aproove for custom reporting. Crystal Reports custom reports can call the SearchJobs function with an AQL query string, returning the matching project list as the report's input set (TN 60).