Auto-Sync RAG Pipeline: Blob Events to Azure AI Search with Terraform ⚡

Azure AI Search indexers can run on a schedule, but that means stale results between intervals. Here's how to build an event-driven pipeline with Event Grid, Service Bus batching, and a Function App to trigger indexer runs on demand using Terraform. You've built your Azure AI Search index, configured a blob indexer, and set it to run on a 5-minute schedule. But that means documents uploaded between intervals sit unindexed. For internal tools this might be fine. For customer-facing RAG, 5 minutes of stale results is too long. Azure AI Search indexers support scheduled runs (minimum every 5 minutes) or on-demand triggers via the REST API. This post builds an event-driven pipeline with Terraform: Blob storage events fire through Event Grid, batch in a Service Bus queue, and invoke a Function App that calls the Run Indexer API. Near-real-time sync without polling. 🎯 🏗️ Architecture Overview Blob Storage (docs added/updated/deleted) ↓ Event Grid Subscription Service Bus Queue (batches event