
Audit-ready, any day. That’s the promise of continuous compliance, and it starts with one change: every access review arrives with the answer.
Access Certifications is Oak’s identity-agnostic access review. It reviews people, service accounts and AI agents across Active Directory, SAP, cloud and SaaS in a single campaign, and recommends keep, revoke or downsize on every row before a reviewer opens it.
Watch how Access Certifications reviews every identity in one campaign.
The review nobody trusts
Every enterprise runs one. Call it the quarterly SOX review of the financial systems: a spreadsheet exported from SAP and the directory, sent to forty managers with a two-week deadline.
A finance director opens her queue on a Monday. It holds every entitlement her team has on the ERP: a few hundred rows of role codes she can’t read, with no sign of who last used what. She approves them all by Friday, because revoking the wrong one could stop the month-end close.
The service accounts that run the nightly close aren’t in the spreadsheet. Neither is the AI agent the finance operations team added last quarter to reconcile invoices. The two revokes that did come back went to an inbox, and nobody can say whether they happened. When the auditor asks for evidence, the IAM team spends a week rebuilding it from screenshots.
A large estate runs dozens of campaigns like it every year. Access Certifications was built so none of them ends in a rubber stamp.
Every reviewer, one view
Progress, revoked and pending items, and every reviewer’s status in one campaign.
Scope, decide, remediate, prove
Every Oak campaign runs in four moves. All four read from the same graph, where Oak models every identity and what it can really do.
1. Scope: only the access in question
You choose the scope, the reviewers and the schedule. Filter by app, permission, sensitivity, grant type or last use, and set the campaign to repeat.
Then Oak scopes it smaller. Scope comes from the mined role model, so reviewers see the grants that sit outside a person’s role, not every entitlement on the app.
People, service accounts and AI agents land in one queue and one evidence package. An identity without a manager routes to its owner’s manager, so nothing is left unreviewed. In the SOX review, the batch accounts and the reconciliation agent now sit beside the finance team, each with a named reviewer.
2. Decide: the answer on every row
Before a reviewer opens the campaign, Oak has done the analysis. Every row shows when the access was last used, what peers hold, how sensitive it is and how it was granted, with a keep, revoke or downsize verdict and the finding behind it. The riskiest rows come first.
Reviewers are notified in Slack and email, and transfer any row that isn’t theirs. The finance director now opens a short queue that starts with a posting right nobody on her team has used in a year. The verdict is revoke, and the reason is on the row. Managers and app owners spend minutes, not afternoons, so the review never disrupts the business.
3. Remediate: carry every decision through
A revoke is a decision, not a result. Each one becomes a required action and a ticket, and reminders and escalation run on their own, so nobody chases reviewers by email.
Oak never removes access itself: your team applies every change. Then Oak re-reads the graph and reports which changes took effect. The two revokes that used to vanish into an inbox now close with proof.
4. Prove: and keep proving it
Export the full decision log: immutable evidence of who decided what and why, with proof the list was complete. The auditor gets it the day they ask.
The graph keeps watching after the campaign closes. Between campaigns, Oak flags access that was certified but never used, or revoked and then regranted, so you know whether the last review held.
Scope, decide, remediate, prove
Every campaign is scoped, decided, carried through and proven from the same graph.
Built on the activity-based identity graph
A review built on grants alone can only ask whether access exists. Oak asks whether it’s needed. No one else has an activity-based identity graph, and the graph that inventories every identity in your estate drives every review.
Verdicts you can trace
Every recommendation points to a named finding: last use, peer access, sensitivity or grant type. A reviewer can question it, and an auditor can follow it.
One graph, every identity
People, service accounts and AI agents share one model, so one campaign covers all of them. No separate tool for non-human identities, and no second review to schedule.
Reviews that stay true
The graph keeps updating after the campaign closes, so a certification is a starting point, not a snapshot. You can see whether the last review held, and act before the next one.
It’s built for scale. At a global US financial services company with 40,000 employees, one campaign covered 86,000 entitlements, with recommendations generated at launch.
{{small-cta}}
What changes for each team
The review stops being a quarterly scramble and becomes a control that runs on its own.
- CISO: access that actually shrinks. Reviews remove real access across every identity type, instead of documenting it.
- GRC: audit-ready, any day. Immutable evidence of who decided and why, with proof the list was complete, ready the day the auditor asks.
- IAM team: no more chasing. Reminders and escalation run on their own, and every change is confirmed from the live graph.
- Reviewers: minutes, not afternoons. Only the access in question, riskiest first, with the answer already on the row.
From rubber stamp to real reduction
Access reviews are where an identity governance program starts, and a step towards an Autonomous Identity Governance Program. As the mined role model sharpens, campaigns get smaller and every review counts.
See it on your own estate. We run a campaign on your applications during the proof of value.
See it, understand it, act on it, with every action on record.
Bring the campaign your reviewers dread most. We’ll run it on your own estate.
Built for giants. Born complete.
Discover related posts


