Prodigy - OpenText Content Storage Service Integration and Automation

Common Integration Use Cases Between Prodigy and OpenText Content Storage Service

Prodigy and OpenText Content Storage Service complement each other well in enterprise AI and content operations. Prodigy handles the active labeling and annotation of training data, while OpenText Content Storage Service provides secure, scalable, compliant object storage for large volumes of source content, intermediate datasets, and annotation outputs. Together, they support efficient machine learning workflows with stronger governance, lower storage overhead, and better collaboration across data science, operations, and compliance teams.

1. Centralized storage for raw training data and annotation inputs

Data flow: OpenText Content Storage Service to Prodigy

Enterprises can store large source datasets such as images, documents, audio files, or video clips in OpenText Content Storage Service and expose them to Prodigy for annotation jobs. This is especially useful for computer vision, document classification, and NLP projects where data volumes are high and need to remain centrally managed.

• Reduces duplicate copies of training data across teams
• Improves access control and retention management for sensitive content
• Allows data scientists to label directly from a governed enterprise repository

2. Persisting labeled datasets for model training and auditability

Data flow: Prodigy to OpenText Content Storage Service

After annotation is completed in Prodigy, labeled datasets can be exported and stored in OpenText Content Storage Service as the authoritative version for downstream model training. This creates a durable record of the exact dataset used for a given model version, which is valuable for reproducibility, audit readiness, and regulated environments.

• Supports traceability between labeling cycles and model releases
• Enables versioned storage of training corpora and label files
• Helps compliance teams retain approved datasets according to policy

3. Active learning loop with storage-backed data refresh

Data flow: Bi-directional

Prodigy?s active learning workflow can be paired with OpenText Content Storage Service to continuously refresh the pool of unlabeled content. New documents or media files can be ingested into OpenText, selected for annotation in Prodigy based on model uncertainty, and then written back as labeled outputs for the next training cycle.

• Speeds up iterative model improvement by prioritizing high-value samples
• Supports continuous learning programs in fraud detection, quality inspection, and content classification
• Keeps the full content lifecycle within enterprise storage governance

4. Secure storage for domain expert review packages

Data flow: Prodigy to OpenText Content Storage Service

Organizations often need subject matter experts to review labeled samples, edge cases, or disagreement sets. Prodigy can export these review packages to OpenText Content Storage Service so business users can access them through controlled enterprise content processes without needing direct access to the annotation environment.

• Improves collaboration between data science teams and business reviewers
• Supports controlled distribution of sensitive content to legal, compliance, or operations teams
• Creates a durable review archive for quality assurance and governance

5. Long-term retention of annotation history and model training evidence

Data flow: Prodigy to OpenText Content Storage Service

For regulated industries such as healthcare, financial services, and public sector organizations, it is important to retain evidence of how AI models were trained. Prodigy annotation logs, label exports, and dataset snapshots can be archived in OpenText Content Storage Service to support internal audits, model risk management, and regulatory inquiries.

• Preserves historical annotation decisions and dataset versions
• Supports governance for AI model validation and approval processes
• Reduces risk associated with undocumented training data changes

6. Scalable storage for large media and document annotation programs

Data flow: OpenText Content Storage Service to Prodigy and Prodigy to OpenText Content Storage Service

Enterprises running large-scale labeling programs for image libraries, scanned documents, or video archives can use OpenText Content Storage Service as the primary repository for all content assets. Prodigy then accesses only the subsets needed for each labeling sprint, and completed annotations are stored back in OpenText for long-term management.

• Supports high-volume annotation initiatives without overloading local systems
• Improves storage efficiency by keeping master content in one cloud repository
• Enables better lifecycle management for raw, in-progress, and completed datasets

7. Migration path from legacy content repositories into AI labeling workflows

Data flow: Legacy content systems to OpenText Content Storage Service to Prodigy

Organizations modernizing legacy storage can migrate unstructured content into OpenText Content Storage Service and then use Prodigy to label selected content for AI use cases. This is a practical approach for enterprises that want to unlock value from archived documents, historical images, or legacy case files.

• Turns dormant content archives into usable AI training assets
• Provides a governed storage layer before data enters the labeling process
• Helps teams prioritize high-value content for model development

Overall, integrating Prodigy with OpenText Content Storage Service helps enterprises build a controlled, scalable AI data pipeline. OpenText provides the secure content foundation, while Prodigy adds the human-in-the-loop labeling capability needed to create high-quality training data efficiently.