Engineering and architecture
Bentley ProjectWise archives: what's inside and what to keep
By SourceX Editorial · Updated
Short answer
A ProjectWise archive holds far more than files: version history, attributes, workflow states and, where enabled, an audit trail of who did what and when. Keep the database and storage areas together, export audit and workflow history before any migration, and exclude public agency and critical infrastructure projects by default from any reuse.
Key takeaways
- A ProjectWise datasource pairs a database of metadata with storage areas that hold the files; one without the other is incomplete.
- Audit trails and version histories are process evidence: they show review cycles, approvals and rework, not just final drawings.
- Migrations often carry current files but leave history behind, so export audit and workflow records before the old system is retired.
- User names in audit trails are personal details, so pseudonymize them consistently before any reuse.
- Public agency and critical infrastructure projects should be excluded by default from any licensing package.
What does a ProjectWise archive actually contain?#
A ProjectWise archive contains a datasource, which pairs a database of metadata with one or more storage areas holding the files. The database is where the history lives: folder and work area structures, document attributes, versions, workflow states, reference relationships, permissions and, where enabled, the audit trail.
Configuration varies by firm and by version, so the first job is an inventory. Ask your ProjectWise administrator which features were switched on and when, because an audit trail that was disabled for years, or purged on a schedule, changes what the archive can show.
- Documents in storage areas, including DGN, DWG, PDF and Office files.
- Version history for documents that were versioned rather than overwritten.
- Attributes and environments: discipline, sheet number, status, revision and custom fields.
- Workflow and state definitions, such as work in progress, check, approved and issued.
- Audit trail events, such as creation, check-out, check-in, state changes, moves and deletions.
- Reference attachment relationships between design files.
- User, group and access control lists.
Why audit trails and versions count as process evidence#
Audit trails and version histories count as process evidence because they record how a design moved from draft to issue, not just where it ended up. A run of check-outs, state changes and new versions around a review date shows who checked a sheet, how much rework followed and how often a deliverable bounced back.
AI developers value records that link a request to a decision and an outcome. In a ProjectWise datasource, that link appears when workflow states, review markups and versions line up: a sheet moves to check, a reviewer's markup is attached, the designer checks in a revised version, and the sheet moves to approved.
| ProjectWise element | What it records | Value as process evidence | Default handling |
|---|---|---|---|
| Audit trail | Who acted on which document and when | High: shows review and rework sequences | Keep; pseudonymize user names |
| Versions | Prior states of a document | High when paired with review markups | Keep with the audit trail |
| Workflow states | Draft, check, approved and issued transitions | High: the approval path in structured form | Keep definitions and history |
| Attributes | Discipline, sheet, revision and status fields | Medium: makes records searchable and linkable | Keep; review custom fields for client data |
| Reference relationships | Which files reference which | Medium: shows how models are assembled | Keep for internal use; review before reuse |
| Final issued files | Drawings and specifications | Low on their own; client-facing | Retain; usually excluded from licensing |
- 01Your work records
- 02Personal details removed
- 03You approve the release
- 04Licensed to AI developers
What to keep when you migrate or retire ProjectWise#
When you migrate or retire ProjectWise, keep the database and storage areas together as a complete, restorable archive before anything moves. Migration tools often carry current files and selected attributes to the new platform while leaving versions, audit history and workflow records behind.
Treat the export as a records project, not an IT ticket. General counsel and practice leaders should decide which history matters before the old servers are switched off, because a deleted audit trail cannot be rebuilt from the new system.
- Back up the datasource database and every storage area at the same point in time.
- Export the audit trail to a structured format, with document identifiers that match the file export.
- Export workflow, state and environment definitions so attribute values can be read later.
- Save a user and group mapping, then decide how names will be pseudonymized for any reuse.
- Record which projects, by client type, sit in which folders or work areas.
- Keep the ProjectWise version and configuration notes with the archive.
Which projects should be excluded by default?#
Projects for public agencies and critical infrastructure should be excluded by default. Civil and infrastructure firms often run ProjectWise because transportation departments and utilities expect it, and those clients' agreements commonly restrict reuse, require return of records or treat some information as security-sensitive.
Default exclusion is a practical rule, not a legal conclusion. A later review may clear specific private work in the same datasource, but starting from exclusion keeps utility systems, bridges, transit facilities and energy assets out of any package unless counsel and the firm decide otherwise.
| Project type | Default | Why |
|---|---|---|
| State DOT and transit work | Exclude | Agency terms, public records rules and security sensitivity |
| Water, wastewater and energy utilities | Exclude | Critical infrastructure and agency restrictions |
| Federal and military facilities | Exclude | Security and export-control questions |
| Private land development | Review | Often on firm terms; check confidentiality |
| Private industrial and commercial work | Review | Check client confidentiality and process data |
| Internal standards, seed files and templates | Candidate | Firm-built content with no client data |
Personal details hiding in ProjectWise#
Personal details in ProjectWise sit mostly in the metadata. Audit trail entries carry user names and sometimes machine names, attributes may record checker and approver names, and documents filed from email can include signatures and phone numbers.
Pseudonymize user identifiers consistently, so one person maps to one token across the whole archive. That keeps the review sequence intact while removing who did the work. Professional seals and signatures on issued sheets are another reason issued drawings usually stay out of scope, and employees deserve a plain explanation of how their work records will be used before any package is approved.
Illustrative: a civil firm moves off on-premises ProjectWise#
Illustrative: a fictional civil engineering firm with transportation, water and private land development practices is moving from an on-premises ProjectWise datasource to a cloud document platform. The migration plan carried current files only.
Its digital practice leader held the cutover long enough to back up the datasource and storage areas together and to export the audit trail and workflow history. Transportation and water work was tagged as excluded by default. Private land development work on the firm's own terms became the candidate set: audit trail sequences, workflow transitions and internal review markups, with user names pseudonymized and project identifiers removed. Issued drawings stayed out.
How SourceX approaches ProjectWise archives#
SourceX approaches ProjectWise archives as process records first and drawings second. During Supply, the firm describes its datasources, record families and project types using metadata only; nothing is shared during the initial assessment. Rights and Preparation then apply the default exclusions, pseudonymize users and strip client identifiers.
Large archives stay in the firm's own storage or ship on encrypted drives, and SourceX never hosts multi-TB datasets. The approved scope, privacy record and release authorization are documented in a SourceX Evidence Packet.
Frequently asked questions
Can we keep ProjectWise read-only instead of exporting?
Some firms do, but a read-only server still needs maintenance, security patching and someone who can restore it years later. A complete backup plus structured exports of audit and workflow history is easier to keep usable. Check your Bentley agreement for what access continues if the subscription ends.
How do we know whether our audit trail is complete?
Do not assume it is. Audit trail settings, retention and purging are configured by administrators, and many firms changed them over time. Ask the administrator which events were logged and whether old records were purged, and note any gaps in the inventory before relying on the history.
Do client-hosted ProjectWise datasources belong to us?
Usually not. Some agencies host their own datasource and give consultants access. Records there follow the client's terms and should be treated as client-controlled. Only datasources the firm operates under its own license are candidates for review.
Are seed files and CAD standards worth keeping?
Yes, for operations, and they are often firm-built. Seed files, cell libraries and level standards describe how the firm works. Some are adapted from agency CAD manuals, though, so check which parts came from a client before treating them as the firm's own.
Should review markups kept outside ProjectWise be archived too?
Yes, if review happened in a separate markup tool. Many firms mark up PDFs in Bluebeam and file only some of them back into the datasource. Without the markups, workflow states show that a review happened but not what the reviewer said, so capture both and link them by document identifier.
What if documents were overwritten rather than versioned?
Then the file history is thin, but the audit trail may still show check-out, check-in and state change events. That sequence is useful evidence on its own. Record the gap in the inventory so no one overstates what the archive contains.
Related resources
- IndustryHealthcare data
- IndustryConstruction data
- QuestionDo AI companies buy private business data?
- InsightCan construction companies sell their data to AI companies?
- InsightCan restaurants and hospitality groups sell their data to AI companies?
- InsightCan SaaS companies sell their data to AI companies?
See if your company qualifies
A short company assessment. No data uploads are needed.