@Sabeek Ahmed

Yes, you can do this. First use Azure Video Indexer to analyze the video and produce JSON with transcript, topics, entities, sentiments, and flags. Next either expose a Pega Service REST to receive a webhook when the analysis is ready or poll with a Pega REST connector triggered by a queue processor. Create a Pega Data Type to store the fields you need like topics, moderation flags, confidence, and the video id. In the case flow add a step that ingests the JSON and maps it with a Data Transform. Build a Pega GenAI prompt template that takes those fields and asks the model to return a routing label, priority, and a short reason using your business rules. Set case properties from the model output such as RouteTo work queue, work group, urgency, and rationale. Add guardrails with confidence thresholds, rule based overrides, and a manual review path for uncertain cases. Log prompt, model output, model version, and source facts for audit and mask sensitive text before sending to the model. Run ingestion and GenAI steps asynchronously with a queue processor, add retries, and watch Pega alerts for failures.