Prodigy - OpenText Information Archive Integration and Automation

Common Integration Use Cases Between Prodigy and OpenText InfoArchive

1. Archival Data Sampling for AI Model Training

Data flow: OpenText InfoArchive ? Prodigy

Organizations can extract representative samples from archived customer records, claims, invoices, emails, or case files stored in OpenText InfoArchive and send them to Prodigy for annotation. This is useful when building NLP or document classification models that need historical, compliant data for training.

• Improves model accuracy by using real enterprise records rather than synthetic data
• Supports controlled access to archived content for data science teams
• Reduces manual effort in locating and preparing training datasets

2. Compliance-Safe Labeling of Sensitive Historical Content

Data flow: OpenText InfoArchive ? Prodigy

When teams need to label archived documents containing regulated or sensitive information, InfoArchive can provide governed access to the source content while Prodigy handles the annotation workflow. This enables legal, compliance, and AI teams to collaborate without moving data into unmanaged repositories.

• Maintains retention and access controls during annotation
• Supports privacy-aware review of legacy records
• Helps organizations build AI models without violating data governance policies

3. AI-Assisted Classification of Archived Records

Data flow: Prodigy ? OpenText InfoArchive

After training a classification model in Prodigy, the resulting labels or prediction outputs can be written back to OpenText InfoArchive as metadata. This is valuable for categorizing archived content by document type, retention class, legal hold status, or business function.

• Improves search and retrieval across large archives
• Automates metadata enrichment for legacy content
• Supports more accurate retention and disposition decisions

4. Active Learning on Archived Case Files and Correspondence

Data flow: OpenText InfoArchive ? Prodigy ? OpenText InfoArchive

InfoArchive can supply a large archive of case files, correspondence, or transaction records to Prodigy, where active learning identifies the most informative items for labeling. Once labels are confirmed, they can be stored back in InfoArchive to create a governed record of classification outcomes.

• Minimizes labeling effort by focusing on high-value samples
• Creates a repeatable workflow for large-scale document classification
• Preserves an audit trail of AI-assisted decisions

5. Legacy System Decommissioning with AI Readiness

Data flow: Legacy system ? OpenText InfoArchive ? Prodigy

During legacy application retirement, data is moved into OpenText InfoArchive for compliant long-term retention. Selected datasets can then be exported to Prodigy to train models that automate future classification, extraction, or routing tasks previously handled by the retired system.

• Reduces dependency on obsolete applications
• Preserves access to historical data for model development
• Supports modernization programs with both archiving and AI enablement

6. Training Data Preparation for Document Extraction Models

Data flow: OpenText InfoArchive ? Prodigy

Archived forms, contracts, invoices, and correspondence can be retrieved from InfoArchive and annotated in Prodigy to train extraction models for fields such as names, dates, amounts, and reference numbers. This is especially useful for automating document processing in finance, HR, and operations.

• Accelerates creation of structured datasets from unstructured archives
• Improves downstream automation for document processing
• Enables reuse of historical records for model training

7. Governance of AI Training Data and Label History

Data flow: Prodigy ? OpenText InfoArchive

Organizations can archive annotation outputs, labeling decisions, and training dataset versions from Prodigy into OpenText InfoArchive to meet audit, governance, and retention requirements. This is useful for regulated industries that need to demonstrate how AI models were trained and validated.

• Preserves evidence of data preparation and labeling decisions
• Supports auditability for model development processes
• Helps satisfy internal and external compliance reviews

8. Cross-Functional Review of Archived Content for AI Projects

Data flow: Bi-directional

Business users, compliance teams, and data scientists can use InfoArchive as the controlled source of archived content while Prodigy provides the collaborative labeling environment. Review outcomes, exceptions, and metadata updates can be synchronized back to InfoArchive to keep the archive aligned with AI project findings.

• Enables structured collaboration between legal, records management, and AI teams
• Improves consistency in labeling archived enterprise content
• Creates a governed workflow from archive retrieval to model training and back