Microsoft
Microsoftpablocastro I created a Sharepoint Indexer and Sharepoint Index and I would like ChatGPT to retrieve data from any index within this Search Service. How does one go about modifying that. Currently, it's just retrieving data from the index generated from the demo. Can I make these changes in the UI or do I need to re-deploy code.
UPDATE: After examining the backend code, it appears I can only use one index but a work-around from this https://learn.microsoft.com/en-us/rest/api/searchservice/indexer-operations article would allow multiple indexers writing into the same index. I I pointed my Sharepoint Indexer to use the current index and RESET the index before running . (Interesting for the Sharepoint Datasource, I had to re-run API rest code to generate a new Verification code because code had expired).
I noticed the content returned for the Sharepoint was the entire document, which could span 4-5 pages of pdf. When prompting for this in the Chat window, it exceeded 16385. What is the best practice for this scenario. Would you be able to point me to where I can chunk the SharePoint content to allow limiting content we send to Azure Open AI due to token limits?
