I'm building a Agent on OpenAI that I can give access to a vector store in which I have multiple files. One file is a proofing report, the next is a inventory report, the next a knowledge file where I tell it how to use the other reports, and finally a refrence report which I explain is to look at the format of the response I want. My problem is I need the agent to look at the proofing file and analyze it, then create a list of the items that are missing a large amount or concerning, next I need it to look at the inventory report and find the items one by one all of their locations, the amount of bottles and other things like that from that file. Finally, I need it to generate a final list of all the items it had from the first list with their locations, bottle size and a couple of other information. Long story short no matter what prompt I use or what agent, it doesn't seem to output results like I need it to. The data Im using is alot so I'm forced to use models like 4o mini instead of the regular 4o. I don't know how to go about it, I'm working with a huge company since I know the owner and don't want to say I can't do it but I'm genuinely stuck. If anyone has developer experience or has worked on things like this for a while please help me out.