Share
Report
1Hello, I'm really new to this.
I have an on-prem (ASP.NET) intranet application that allows users to search our catalog of current and past projects. The data is currently stored in an XML file and describes a few thousand projects. Each project has information such as a project number, name, client, project manager, and status as well as a list of project phases, each of which has a phaseID, phase name, and status. The current app lets users search for projects based on project numbers or text contained in any of the other fields. For example, they might search for all projects where "John Doe" is the project manage and "Sacramento" appears in any of the project or phase names.
I'd love to migrate this to Copilot but I'm not sure how. The XML file itself is bigger than the upload limit, and when I created a simple copilot where I uploaded just a small subset of the XML file, Copilot clearly couldn't make any sense of the XML.
I was thinking that one option would be to convert the XML file to an individual PDF for each project which followed a predictable format. I set up a half dozen such PDF files and then created a new model which I trained to extract the data from the files. This seems to get the correct data. I've uploaded a sample of the PDF.
What I was hoping to do was to upload all of the PDFs to a SharePoint folder and create a Copilot that would use only the files in that folder, with help from the model I created, to handle prompts like "Show me all projects managed by Groucho that have a Design phase."
Am I on the right path?
All responses (
Answers (
