Usage Limits for a Public App with High OpenAI API Interactions

We are in the process of developing a public application that involves a significant number of interactions with the OpenAI API. However, during the development and testing phase, we found ourselves reaching the usage limit within just four days. Understandably, we will need to upgrade to a package with more allowances, but we are concerned that even with this change, our limits may still be insufficient once the application is launched. This is mainly due to the nature of our app, which heavily relies on updating tables. While we are working on optimizing this aspect through relational improvements, we remain apprehensive that our current limitations may continue to pose challenges during development and beyond.

I assume you’re using the native OpenAI integration and those do consume updates. How many updates have you consumed so far? Do you expect it to scale up a lot when it reaches production?