OCR and document extraction

Combine native extraction and OCR before validating structured records.

ingest documents→OCR pagesextract native text→classify sections→extract fields→validate schema→publish records

OCR and document extraction

Activity responsibilities

Activity	Responsibility
ingest-documents	Execute the ingest documents stage and publish its declared outputs for downstream activities.
ocr-pages	Execute the OCR pages stage and publish its declared outputs for downstream activities.
extract-native-text	Execute the extract native text stage and publish its declared outputs for downstream activities.
classify-sections	Execute the classify sections stage and publish its declared outputs for downstream activities.
extract-fields	Execute the extract fields stage and publish its declared outputs for downstream activities.
validate-schema	Execute the validate schema stage and publish its declared outputs for downstream activities.
publish-records	Execute the publish records stage and publish its declared outputs for downstream activities.

Inputs

• Documents
• Extraction schema
• OCR model

Outputs

• Page text
• Structured records
• Validation report

Provenance to preserve

• Page coordinates
• Confidence
• Validation failures

Runnable example

This is the complete checked-in bundle for this pattern. Download the environment, scope, topology, workflow, input, container recipe, runner, and validation contract from this page before executing it.

Read setup instructions ↓Environment YAML ↓Execution scope YAML ↓Network topology YAML ↓Workflow YAML ↓Input data ↓Container recipe ↓Run the workflow ↓Validate outputs ↓Expected output manifest ↓

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

Open verified output manifest ↓

Execution considerations

Route pages according to text-layer quality and retain coordinates linking every field to source evidence.

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