Thing I Want to Invest In: Scalable Streamlit

You get 10 lines of code and one CLI command to create and deploy an application for 100 users. Actually, let's make it 1,000.

What? Building a scalable back-end solution for hosting Streamlit applications outside of Snowflake.

Rivers of data as a metaphor.

Why? I love Streamlit for building data applications in pure Python. I've never liked structured low-or-no-code. Jupyter Notebooks are great for analysis, but don't produce an "application" other users can interact with. HoloViz is arcane. After you've had every Python library at your disposal, BI tools feel crushingly restrictive.

There's a big problem, though.

Streamlit is great for generating a frontend and basic hosting mechanism for a few users. Want 50+ concurrent users? You have some work to do! The cacheing mechanism is clever, but only goes so far.

I saw this coming when Snowflake bought Streamlit. Indeed, if you are in the Snowflake ecosystem, they have solutions available. That creates a huge opportunity to service... everyone else!

Snowflake won't do it, but someone else can.

Vision. You get 10 lines of code and one CLI command to deploy a Streamlit (or similar) application on the customer's infrastructure for 100 concurrent users. Let's call it "10x".

Doing that? reachme@jtouellette.com

Subscribe to Joshua T Ouellette

Don’t miss out on the latest issues. Sign up now to get access to the library of members-only issues.
jamie@example.com
Subscribe