My use case is simple, I want an OpenAI Assistant and train it on my companies policies (annual leave, maternity, etc…) and IT Support FAQ’s for employees.
The users can then ask the AI via a service I make to ask the AI questions and get relevant answers based on the docs uploaded.
I can’t seem to figure out how much this will cost me, or if this is the best option or the most cost efficient option.
Please help!
Welcome to the community!
Unfortunately, the only way to discover that is through good old value engineering. Where did you get stuck?
If your employees are already using GPT plus, teams, or enterprise, and the data isn’t super sentitive then the best bet would probably be a custom gpt, as opposed to an assistant.
If you need certifiably high accuracy, you will probably need a properly engineered rag system.
Thanks @Diet - How would I create a rag system? Are there any guides?
You’d need some coding. Here’s a guide for making one.