Prodigy - OpenText Internet of Things Platform Integration and Automation

Common Integration Use Cases Between Prodigy and OpenText Internet of Things Platform

1. Sensor Data Labeling for Predictive Maintenance Models

Data flow: OpenText Internet of Things Platform ? Prodigy

OpenText Internet of Things Platform ingests machine sensor readings, vibration patterns, temperature spikes, and equipment status events from manufacturing assets. Relevant time windows and event segments are exported to Prodigy for expert labeling, such as ?normal operation,? ?bearing wear,? ?motor anomaly,? or ?failure precursor.? Data scientists then use the labeled dataset to train predictive maintenance models that identify early warning signs before equipment breakdowns.

Business value: Reduces unplanned downtime, improves maintenance planning, and helps maintenance teams focus on high-risk assets instead of reacting to failures.

2. Quality Inspection Model Training from Connected Production Lines

Data flow: OpenText Internet of Things Platform ? Prodigy

In smart manufacturing environments, OpenText Internet of Things Platform collects images, sensor signals, and inspection triggers from production equipment and connected cameras. Prodigy is used to label defect images, classify product anomalies, and annotate inspection outcomes such as scratch, misalignment, contamination, or pass/fail. These labeled samples support computer vision models used for automated quality control.

Business value: Improves defect detection accuracy, reduces manual inspection effort, and supports consistent product quality across production lines.

3. Root Cause Analysis Dataset Creation from IoT Alerts

Data flow: OpenText Internet of Things Platform ? Prodigy

When OpenText Internet of Things Platform detects abnormal sensor patterns or generates operational alerts, those incidents can be exported into Prodigy for expert review and labeling. Engineers can tag the likely cause, such as calibration drift, overheating, network interruption, or operator error. The resulting labeled incidents create a structured dataset for training classification models that support root cause analysis.

Business value: Speeds up troubleshooting, improves incident triage, and helps operations teams identify recurring failure patterns more quickly.

4. Active Learning Loop for Edge and IoT Anomaly Detection

Data flow: OpenText Internet of Things Platform ? Prodigy ? OpenText Internet of Things Platform

OpenText Internet of Things Platform streams large volumes of sensor data to an analytics pipeline. A preliminary anomaly detection model flags uncertain or borderline cases and sends them to Prodigy for human labeling. Prodigy?s active learning workflow prioritizes the most informative samples, allowing domain experts to label only the most valuable events. The updated labels are then fed back into the IoT analytics environment to retrain and improve anomaly detection models.

Business value: Minimizes labeling effort, improves model performance faster, and creates a continuous improvement loop for operational analytics.

5. Event Classification for Logistics and Fleet Operations

Data flow: OpenText Internet of Things Platform ? Prodigy

For logistics organizations, OpenText Internet of Things Platform collects telematics data from vehicles, trailers, and shipping assets, including GPS, temperature, shock, and door-open events. Prodigy can be used to label event sequences such as route deviation, cold-chain breach, unauthorized access, or delivery exception. These labels help train models that classify shipment risk and support exception management workflows.

Business value: Improves shipment visibility, reduces spoilage and loss, and enables faster response to delivery exceptions.

6. Training NLP Models on IoT Service and Maintenance Notes

Data flow: OpenText Internet of Things Platform ? Prodigy

OpenText Internet of Things Platform often integrates with enterprise systems that generate maintenance logs, technician notes, and incident summaries tied to device events. These text records can be exported to Prodigy for annotation, such as issue type, severity, asset category, or recommended action. The labeled text is then used to train NLP models that automate ticket routing, summarize incidents, or extract maintenance insights.

Business value: Reduces manual case handling, improves service desk efficiency, and turns unstructured maintenance text into usable operational intelligence.

7. Closed-Loop Model Improvement for Connected Operations

Data flow: Bi-directional

OpenText Internet of Things Platform supplies real-world device and sensor data to Prodigy for annotation, while Prodigy returns validated labels and training outputs to the IoT analytics environment. This supports a closed-loop workflow where operational teams, data scientists, and subject matter experts continuously refine models for anomaly detection, classification, and forecasting. The integration can also support periodic retraining based on new device behavior or seasonal operating conditions.

Business value: Keeps models aligned with changing field conditions, improves long-term accuracy, and supports scalable AI operations across connected assets.