Quick Summary: Oleria Trustfusion, an AI-native identity security & governance platform, delivers AI-Powered Identity Governance for JML and Access Reviews — automating joiner bundles, surfacing three-signal review recommendations, and answering any access decision question through Copilot. IAM engineers move faster and audit trails capture what operators saw when they decided.
Governance decisions — who gets access at joiner, what changes at mover, who keeps it at review, what a one-off request justifies — are judgment calls. The judgment combines role context, peer comparison, usage history, recent changes, and risk classification. Doing that judgment manually for hundreds of decisions a week is the bottleneck; doing it without context is the cause of audit findings.
AI-assisted governance has to be transparent and grounded. Black-box "approve this," opaque scoring, or AI that decides without explanation is worse than no AI. The AI's job is to surface the right context per decision and recommend; the human decides. Today this lives across three governance workflows; tomorrow it surfaces inline at the decision point.
Across the governance lifecycle — joiner bundles, access reviews, access decisions — Oleria's AI surfaces the trade-offs from real graph and activity data. The operator decides. In-workflow surfacing in the request/review UI plus AI-context capture in the audit are coming next; mover (the third JML stage) lands on the same surface.
Joiner bundle compute time Manual mapping → minutes today
Access review per-line decision time Reduced by recommendation rating today
Time to context for ad-hoc access decisions Seconds today (via Copilot)
In-workflow AI context + audit capture Coming
Joiner: peer-attribute matching computes the new hire's recommended bundle from observed usage of identities sharing the same job attributes (title, department, location). The IAM engineer reviews; the bundle reflects what role-holders actually use, not what a static template said. Mover (the next JML stage) is coming with multi-match union bundle resolution; leaver workflows are operational.
Three signals on every access line: dormancy (last-activity-in-days), peer-group coverage (what fraction of peers hold this access), HR change (recent role/department/manager change). Combined into a per-line High/Medium/Low recommendation rating. Reviewers bulk-accept the High-confidence matches and spend their attention on the outliers — the lines that need a real decision.
Copilot answers in plain English, in chat. Ask for any context the decision needs — peer comparison, usage of similar permissions, role alignment, recent changes, risk classification — and the AI surfaces it from the access graph and activity stream. Verified data, not generated. The IAM engineer brings the answer back to the decision; the operator decides.
Three things ship on the same path: in-workflow AI context inside the access-request and access-review UI (so the engineer doesn't context-switch to chat for routine decisions), AI-context summary captured in the audit alongside the decision (so the auditor sees what the operator saw), and Mover joining the JML AI surface with multi-match bundle resolution.
The operator. Always. Oleria's AI surfaces context, recommends, and captures the recommendation. The decision and the operator are captured in the audit; once in-workflow context capture ships, the AI-provided context summary is captured alongside, so the audit shows what the operator saw when they decided. The AI is decision support, not authority.
