Azure Cognitive Search and LangChain: A Seamless Integration for Enhanced Vector Search Capabilities

Hello Daniel Gil, yes, it is valid with any index, no matter if you pushed the data directly to it or using any built-in connector or partner connector: Data sources gallery - Azure Cognitive Search | Microsoft Learn. If you're using any built-in connector, including the ones in preview, if you'd like to vectorize the data through the skillset through the skillset pipeline (Skillset concepts - Azure Cognitive Search | Microsoft Learn), you can check this example: https://github.com/Azure/cognitive-search-vector-pr/blob/main/demo-python/code/azure-search-vector-ingestion-python-sample.ipynb. This is an end to end sample, that works as a workaround while chunking and vectorization is supported natively as skills (soon) as part of the pipeline.