Help on designing the app

[x31]

Hi

I have trained LLM, and I want to develop a chat app around it [via Streamlit. I want to keep the data and code of the app private. [for future ability to monetize]

Few questions while I am getting started:

1. I am receiving errors around using the chat_input function, which seems to be related to versioning. Most recent version I am able to get is 1.23.1, is the version 1.24.0 available via pip? Any creative ideas around being able to upgrade the package?
2. I want to use Stripe for payment integration. Is there any existing integration I can use?
3. Any thoughts around integration with an SQL database to build logic around subscription and create feedback for the LLM?

I would really appreciate the support of the community on this.

[momos]

Hi there!

1. Regarding the chat_input function, you can try and run pip install --upgrade streamlit to update your existing streamlit version to the latest one. I don’t think you can find any workarounds for this, since a version >1.24.0 is required to use the chat_input function.

2. Regarding the Stripe integration, I don’t think any pre-built integration exists currently. In the meantime, you can probably connect to the Stripe API for payments. Check out these links for reference:

Stripe API Integration

Has anyone considered integrating their app with Stripe to create a micro service of sorts? I’ve been able to accomplish this with Square’s Checkout API but vastly prefer Stripe’s ecosystem. Update: I was able to create a hosted invoice page from Stripe using their Invoices feature. Not perfect, but it works https://stripe.com/docs/invoicing/hosted

Stripe API Reference

3. An SQL connection does exist in Streamlit. Please check out the following:
   https://docs.streamlit.io/library/api-reference/connections/st.experimental_connection
   Connecting to data - Streamlit Docs

Cheers,
Moiz

[x31]

Thank you for explaining. Do you have any solution to keeping the data and code for the app private?

[momos]

Well, I’m not sure what exactly you mean. What parts of the data or code do you need to be private? For keeping the entire data and the code private, you can just create a private Github repository (the code will only be visible to you). You can then connect your private repo to the server where you want to deploy the app (like Streamlit cloud), and everything should be private.