Looking for feedback/guidance on developing a OpenAI App.
Our use case/ user flow:
Tool/App is triggered and shows a carousel of cards (possible plans for the user)
User click on a card - opens full screen mode
User has the option to personalise the plan
For personalisation we need ~5 more inputs (i.e. age, gender, etc.)
Question: Should these inputs be handled within the app (probably easier for user, def. better for us/app) OR should we prompt the user in enter the requested data in the ChatGPT thread?
OpenAI’s guidance is clear on conversational flow (makes sense). They also mention under poor use cases:
“Require complex multi-step workflows that exceed the inline or fullscreen display modes.”.
It doesn’t seem to be ideal for the user to continue the conversation to input data once in fullscreen mode with easy/fast option to fill form.
Keen to hear the feedback and thoughts of this great community!
I don’t think it’s that bad of an idea to present the user with some input fields in your user flow. Depending on how you’ve crafted the experience, it might not be too heavy.
Even better if you could infer & auto-fill them based on what you know about the user from the context of a conversation (is that possible?). For example, it’d be nice to fill in the age and gender of the user if they are already known so they don’t need to fill it in again.
My rationale is that OpenAI also doesn’t seem to be too prescriptive on what exactly constitutes a “complex multi-step workflow” (for now) from what I can see from the docs.
I like these 2 sentences in the official guide: “Apps are most valuable when they help people accomplish meaningful tasks directly within ChatGPT, without breaking the conversational flow. The goal is to design experiences that feel consistent, useful, and trustworthy while extending ChatGPT in ways that add real value.” I don’t really think you’re breaking this in principle, from what you’re describing
