Input
What you provide
- Participant policies, local datasets, and the privacy rules that bind them.
- Eligibility, aggregation integrity, and privacy bounds.
AuraOne / Products
Loading the product.
Loading capabilities, workflow, and program details.
AuraOne / Products
Loading capabilities, workflow, and program details.
Models / Federated Learning
Follow participant eligibility, the aggregation policy, and the privacy checks. Data boundaries hold.
Model and data teams running distributed training across controlled participants. Each program is configured around the workflow, review policy, and evidence your team needs.
Input
Work
Output
Move from intake to review and handoff with clear owners at every step.
01
Participant policies, local datasets, and the privacy rules that bind them.
Participant attestations
02
Validate the participants. Train locally. Aggregate and attest.
Aggregation record
03
Eligibility, aggregation integrity, and privacy bounds. The training owner, a privacy reviewer, and the model approver.
Named review and decision history
04
A governed model update with participant and privacy evidence.
Privacy review
What you receive
| Outcome | Work | What you receive | Program fit |
|---|---|---|---|
| Advance | The required Federated Learning checks and review are complete. | Participant attestations, Aggregation record, Privacy review | Program fit depends on the source and its version. It also depends on criteria, reviewers, and operating scope. |
| Hold | A required check, reviewer, or approval is missing. Or a source record is unresolved. | What is blocking, who owns it, and the recovery it needs. | The responsible owner and recovery step stay visible until the issue is resolved. |
| Return for work | The source object or workflow requires correction and another review. | The returned items, their owner, and the review route back. | The updated work returns through the same review path before a new decision. |
Inspect the workflow