Prodigy - OpenText Extended ECM - Content Server Integration and Automation

Common Integration Use Cases Between Prodigy and OpenText Extended ECM - Content Server

1. Controlled sourcing of training content from enterprise repositories

Flow: OpenText Extended ECM - Content Server ? Prodigy

Enterprise documents, images, forms, and records stored in Content Server can be selectively exported to Prodigy for annotation and model training. This is useful when AI teams need to label approved business content such as invoices, claims, contracts, inspection photos, or customer correspondence without copying unmanaged files into local folders.

Business value: Keeps training data aligned with governed source content, reduces manual file handling, and ensures only authorized datasets are used for AI development.

2. Annotated content and labels stored back as governed records

Flow: Prodigy ? OpenText Extended ECM - Content Server

After annotation, labeled datasets, review outputs, and labeling decisions can be written back to Content Server as controlled records with metadata such as model version, label schema, reviewer, and approval status. This creates an auditable history of how training data was created and approved.

Business value: Supports compliance, traceability, and reuse of curated datasets across future AI projects.

3. Human review workflow for disputed or low-confidence labels

Flow: Prodigy ? OpenText Extended ECM - Content Server

When Prodigy flags uncertain annotations or edge cases, the item can be routed into Content Server workflow for business review by legal, compliance, quality, or operations teams. Once approved or corrected, the final decision can be returned to Prodigy to update the training set.

Business value: Improves label quality for sensitive use cases such as regulated documents, policy classification, or customer communications while keeping decision-making within enterprise governance processes.

4. Active learning dataset governance and version control

Flow: Prodigy ? OpenText Extended ECM - Content Server

Prodigy?s active learning process can generate successive dataset versions as the model identifies the most valuable samples to label next. Each dataset version, label taxonomy, and sampling rule can be stored in Content Server with retention and version control policies.

Business value: Gives data science teams a controlled way to manage dataset evolution, compare training iterations, and meet audit requirements for model development.

5. Secure access to sensitive content for AI training

Flow: OpenText Extended ECM - Content Server ? Prodigy

Content Server can act as the secure access layer for sensitive content such as HR files, legal documents, or customer records. Prodigy can retrieve only the subset of content that a user is authorized to label, using metadata filters, permissions, and classification rules from Content Server.

Business value: Prevents overexposure of confidential information, enforces least-privilege access, and enables AI training on sensitive enterprise data without weakening governance.

6. Metadata enrichment for labeled datasets

Flow: Prodigy ? OpenText Extended ECM - Content Server

Prodigy can send annotation results back to Content Server as metadata updates, such as document type, entity extraction results, sentiment, defect category, or image classification. These enriched records can then be used by downstream business processes, search, and analytics.

Business value: Turns raw content into structured, searchable enterprise information and improves retrieval, reporting, and process automation.

7. Training data lifecycle management and retention

Flow: Bi-directional

Content Server can manage retention, disposition, and legal hold for source documents and final labeled datasets, while Prodigy handles the active labeling phase. Once a dataset is no longer needed for training, Content Server can enforce retention schedules and disposition rules for both source and derived content.

Business value: Reduces storage sprawl, supports records management obligations, and ensures training data is handled according to enterprise policy.

8. Cross-functional content operations for AI-enabled business processes

Flow: OpenText Extended ECM - Content Server ? Prodigy ? OpenText Extended ECM - Content Server

Business teams can submit content from operational repositories into Prodigy for labeling, while AI teams refine the model using those labels. The resulting predictions or classifications can be stored back in Content Server and used to trigger downstream workflows such as case routing, document indexing, or exception handling.

Business value: Connects content management, AI model training, and operational execution in one governed workflow, reducing manual classification effort and improving process speed.