Prodigy - OpenText Notifications Integration and Automation

Common Integration Use Cases Between Prodigy and OpenText Notifications

1. Annotation Task Assignment and Reviewer Alerts

Data flow: OpenText Notifications ? Prodigy

When new labeling batches are created in Prodigy, OpenText Notifications can alert assigned annotators, reviewers, and team leads that a new task is ready. This is useful for distributed AI teams working across time zones, where timely notification reduces idle time and keeps annotation throughput high.

Business value: Faster task pickup, better workload visibility, and fewer missed assignments.

2. Quality Review Escalations for Low-Confidence Labels

Data flow: Prodigy ? OpenText Notifications

Prodigy can trigger notifications when labels fail validation rules, confidence thresholds drop, or disagreement rates exceed a set limit. OpenText Notifications can route these alerts to quality assurance leads or domain experts for immediate review.

Business value: Improves dataset quality, reduces rework, and helps prevent poor training data from reaching model training pipelines.

3. Active Learning Milestone Updates for ML Teams

Data flow: Prodigy ? OpenText Notifications

As Prodigy selects new samples for annotation based on active learning, it can notify machine learning engineers when key milestones are reached, such as completion of a labeling round, model retraining readiness, or performance improvement thresholds. This keeps data scientists informed without requiring them to monitor the platform continuously.

Business value: Shorter model iteration cycles and better coordination between annotators and ML engineers.

4. Workflow Status Notifications for Annotation Operations

Data flow: Prodigy ? OpenText Notifications

Prodigy can send status updates when annotation jobs move from queued to in progress, under review, approved, or blocked. OpenText Notifications can distribute these updates to operations managers and project stakeholders so they can track progress and intervene when bottlenecks appear.

Business value: Greater operational transparency and improved delivery predictability for AI projects.

5. Exception Alerts for Data Ingestion or Source System Issues

Data flow: Prodigy ? OpenText Notifications

If Prodigy cannot access a data source, encounters malformed records, or detects missing files during ingestion, it can generate an alert through OpenText Notifications to the data engineering team. This is especially valuable when annotation depends on scheduled feeds from enterprise repositories or cloud storage.

Business value: Faster incident response, reduced downtime, and fewer delays in labeling pipelines.

6. Reviewer Approval and Sign-Off Notifications

Data flow: Prodigy ? OpenText Notifications

After a batch of annotations is completed and approved in Prodigy, OpenText Notifications can notify downstream stakeholders such as model owners, compliance reviewers, or release managers that the dataset is ready for the next stage. This supports controlled handoffs in regulated or high-stakes environments.

Business value: Clear approval workflows, better governance, and smoother transition from labeling to model training.

7. Cross-Team Collaboration on Labeling Changes

Data flow: Bi-directional

When label schemas, annotation guidelines, or project settings are updated in Prodigy, OpenText Notifications can inform all impacted teams, including annotators, QA reviewers, and ML engineers. In return, users can acknowledge or request clarification through notification-driven workflows, helping teams stay aligned as labeling requirements evolve.

Business value: Fewer labeling errors, better change adoption, and improved collaboration across data science and operations teams.