diff --git a/articles/search/retrieval-augmented-generation-overview.md b/articles/search/retrieval-augmented-generation-overview.md index b548899849..02aa99dcb1 100644 --- a/articles/search/retrieval-augmented-generation-overview.md +++ b/articles/search/retrieval-augmented-generation-overview.md @@ -70,7 +70,7 @@ The information retrieval system provides the searchable index, query logic, and The LLM receives the original prompt, plus the results from Azure AI Search. The LLM analyzes the results and formulates a response. If the LLM is ChatGPT, the user interaction might consist of multiple conversation turns. An Azure solution most likely uses Azure OpenAI, but there's no hard dependency on this specific service. -Except for features currently in previewAzure AI Search doesn't provide native LLM integration for prompt flows or chat preservation, so you need to write code that handles orchestration and state. You can review demo source ([Azure-Samples/azure-search-openai-demo](https://github.com/Azure-Samples/azure-search-openai-demo)), updated for agentic retrieval, for a blueprint of what a full solution entails. We also recommend [Azure AI Foundry](/azure/ai-foundry/concepts/retrieval-augmented-generation) to create RAG-based Azure AI Search solutions that integrate with LLMs. +Except for features currently in preview, Azure AI Search doesn't provide native LLM integration for prompt flows or chat preservation, so you need to write code that handles orchestration and state. You can review demo source ([Azure-Samples/azure-search-openai-demo](https://github.com/Azure-Samples/azure-search-openai-demo)), updated for agentic retrieval, for a blueprint of what a full solution entails. We also recommend [Azure AI Foundry](/azure/ai-foundry/concepts/retrieval-augmented-generation) to create RAG-based Azure AI Search solutions that integrate with LLMs. ## Searchable content in Azure AI Search