Prodigy - OpenText Extended ECM Platform Integration and Automation

Common Integration Use Cases Between Prodigy and OpenText Extended ECM Platform

1. Controlled training data preparation from governed enterprise content

Data flow: OpenText Extended ECM Platform ? Prodigy

Organizations can use OpenText Extended ECM Platform as the governed source for documents, images, contracts, forms, and other business records, then route selected content into Prodigy for annotation. This is useful when AI teams need training data from approved enterprise repositories rather than unmanaged file shares.

• Business value: ensures only authorized, version-controlled content is used for model training.
• Operational benefit: reduces manual file extraction and duplicate data handling.
• Typical use case: legal, HR, claims, or quality management documents are sampled from ECM and labeled in Prodigy for NLP or document classification models.

2. Human review and labeling of sensitive enterprise documents

Data flow: OpenText Extended ECM Platform ? Prodigy ? OpenText Extended ECM Platform

For high-risk content such as contracts, compliance records, or regulated correspondence, OpenText Extended ECM Platform can store the source documents and final approved annotations, while Prodigy is used for expert labeling. After review, the labeled output can be written back to the ECM repository as a governed artifact.

• Business value: creates an auditable chain from source document to labeled training set.
• Operational benefit: supports compliance teams and subject matter experts working in a controlled workflow.
• Typical use case: tagging clauses, obligations, or risk categories in contract archives for AI-assisted contract analytics.

3. Active learning loop for enterprise document classification

Data flow: Prodigy ? OpenText Extended ECM Platform

Prodigy?s active learning can be used to identify the most informative documents for labeling, while OpenText Extended ECM Platform supplies the broader enterprise content corpus. As the model improves, newly classified documents or confidence scores can be stored in ECM to support downstream business processes such as records management, routing, or search.

• Business value: accelerates model training with fewer labeled examples.
• Operational benefit: improves classification accuracy for large document volumes.
• Typical use case: automatically classifying incoming invoices, customer letters, or policy documents stored in ECM.

4. Annotation of scanned images and visual records for quality or compliance AI

Data flow: OpenText Extended ECM Platform ? Prodigy

When OpenText Extended ECM Platform stores scanned forms, images, inspection records, or visual evidence, those assets can be exported to Prodigy for image labeling. AI teams can annotate defects, document regions, signatures, stamps, or other visual elements to train computer vision models.

• Business value: enables AI use cases on existing archived visual records.
• Operational benefit: centralizes image governance while enabling specialized labeling workflows.
• Typical use case: labeling damaged product images, signed forms, or stamped approvals for automated inspection or verification models.

5. Training data governance and lineage management

Data flow: Prodigy ? OpenText Extended ECM Platform

Once annotation projects are completed in Prodigy, the resulting datasets, label definitions, reviewer notes, and model-training exports can be archived in OpenText Extended ECM Platform. This creates a governed record of what data was used, who labeled it, and which version supported a specific model release.

• Business value: improves auditability and model governance.
• Operational benefit: supports traceability for regulated AI programs.
• Typical use case: storing approved training datasets for finance, healthcare, or public sector AI initiatives.

6. Exception handling workflow for content requiring expert labeling

Data flow: OpenText Extended ECM Platform ? Prodigy ? OpenText Extended ECM Platform

Documents that cannot be confidently classified by rules or automation in OpenText Extended ECM Platform can be escalated to Prodigy for expert annotation. After labeling, the results can be returned to ECM to update metadata, trigger routing, or close the exception case.

• Business value: reduces bottlenecks in content processing operations.
• Operational benefit: combines automated ECM workflows with human-in-the-loop AI labeling.
• Typical use case: ambiguous invoices, unstructured customer correspondence, or exception claims that require manual categorization before processing.

7. Continuous improvement of enterprise search and content understanding

Data flow: OpenText Extended ECM Platform ? Prodigy ? OpenText Extended ECM Platform

Search logs, user corrections, and poorly matched content from OpenText Extended ECM Platform can be exported to Prodigy to create labeled examples for intent detection, entity extraction, or document relevance models. The improved models can then enhance search ranking, metadata extraction, and content recommendations in ECM.

• Business value: improves employee access to relevant information.
• Operational benefit: turns search feedback into structured training data.
• Typical use case: training NLP models to better identify project names, customer entities, or document types in enterprise repositories.

8. Model output enrichment of ECM metadata and workflow decisions

Data flow: Prodigy ? OpenText Extended ECM Platform

After models are trained using Prodigy, inference results such as labels, confidence scores, extracted entities, or classification tags can be pushed into OpenText Extended ECM Platform as metadata. This allows ECM workflows to route content automatically, apply retention rules, or assign review tasks based on AI predictions.

• Business value: increases automation in document-centric processes.
• Operational benefit: reduces manual indexing and routing effort.
• Typical use case: auto-tagging procurement documents, routing HR records, or flagging compliance-sensitive content for review.