Anomaly detection

Train a detector and make threshold selection an explicit workflow decision.

collect signals→build features→train detector→score validation data→choose threshold→publish detector

Anomaly detection

Activity responsibilities

Activity	Responsibility
collect-signals	Execute the collect signals stage and publish its declared outputs for downstream activities.
build-features	Execute the build features stage and publish its declared outputs for downstream activities.
train-detector	Execute the train detector stage and publish its declared outputs for downstream activities.
score-validation-data	Execute the score validation data stage and publish its declared outputs for downstream activities.
choose-threshold	Execute the choose threshold stage and publish its declared outputs for downstream activities.
publish-detector	Execute the publish detector stage and publish its declared outputs for downstream activities.

Inputs

• Signal dataset
• Feature specification
• Operating-cost policy

Outputs

• Detector
• Threshold
• Validation report

Provenance to preserve

• Time window
• Feature code
• Threshold rationale

Source bundle

The executable workflow and its implementation files are maintained in the separate AkôFlow examples repository. The documentation intentionally keeps code, container recipes, generators, and validation scripts out of the narrative page.

Open the complete source bundle ↗

Verified local execution

These captures and the output manifest were produced by the fixture's local Docker run and validator. They are published with the same bundle as the runnable files.

Workflow evidence

Execution evidence

Output evidence

Execution considerations

Preserve time-based splits to prevent leakage. Feature construction may require more data movement than training.

AkôFlow boundary: the engine schedules, deploys, executes, transfers data, and records evidence. The ML or agent framework remains an implementation choice inside each activity.