KEZEL

Workflows that act with permission.

Agent workflows, ETL and scheduled jobs, dispatched to run beside your data. Every step is checked against policy at run time, and every run lands on the in-boundary ledger.

Every run follows your policyA scheduled job or requestCheck permissionsRun locally and record the outcome

Inspect completed and denied actions.

The run ledger records the job, policy decision and outcome for each action. This sample includes a request denied under policy.

Run ledger · sample data

06:00refresh-kyb-queuepolicy okcompleted · 512 cases
06:14etl-fx-ratespolicy okcompleted
07:30score-attention-queuepolicy okcompleted · ranked
09:02export-casesdenied · policysvc-crm

The first workload is usually one of these.

KYC & KYB case workflows

Cases open on a trigger, route on score, and refresh on schedule, beside the core account stores. The banking page walks the full lifecycle.

the banking deployment →

Scheduled refresh & ETL

Jobs that move and shape data inside the boundary. The policy check happens at run time, in the plan itself, instead of living in a scheduler's credential store.

how a plan is checked →

Agent actions with permission

Agent and model-driven steps pass the same run-time policy checks as scheduled jobs. The policy decision and outcome are recorded for review.

where AI models fit →

Add analytics to the workflow deployment.

A first workflow deploys with no other component in the picture. When a dashboard or an engine follows, it attaches to the same runtime and writes to the same ledger.

See KEZEL run a workload like yours.

Tell us which system holds your data and what you need to do with it. We’ll demonstrate on sample data, then scope a pilot with your security team.