I'm working with a client who has a massive collection of data spread across several Azure blob containers. This data came from retired file servers used in past projects. They want to create a web interface that allows users to access these documents on demand and search within them—not just by file names, but diving deeper into their contents. Considering there are millions of files, like Excel, Word, and PDF documents, is it feasible to develop an application that supports this level of searching? I'd also like to incorporate Azure AI to answer queries based on the contents of these files. Has anyone implemented something similar? What approaches do other companies take in such situations, especially for audit functionality?
1 Answer
If you decide to go ahead with it, one route could be to set up a Storage Account with Azure AI Search. You would need to vectorize the content and then use Azure AI Foundry to connect it as a knowledge base. This can help in developing a functional search capability.
Appreciate the insight! That sounds like a solid approach.