Comment by @pealco • Hey

Be careful connecting BigQuery to a public endpoint. You could end up with a big bill.

Stats

Comments

There was some confusion around the data source I used for this because I was in the telegram Lens Data Garden and I shared my post, then asked about connecting to the BigQuery data source immediately after. This is actually using The Graph's free hosted service, which isn't really ideal either because I'm making a bunch of requests as timetravel queries (give me this person's number of followers as of block x, which was closest to midnight of a given day). I'm going to look into the BigQuery data next, Stani had an idea of looking at forecasting collect revenue, and I was going to look into bot activity. If I do make something public, I have an RStudio server that I'd make a custom API on where I can limit the functionality (as in prevent some weird injection attack) and keep things hidden better on a separate server. But to be honest I'd probably just schedule a job that runs hourly that refreshes a public dataset that way the cost is always the same regardless of how many people are using the app since I've done that a lot in the past (this has been running for 2.5 years without changes: https://cryptocurrencyresearch.org/introduction.html ). But tbh it would probably be pretty reasonable to set a budget & several alerts (which I already have for some personal use GPT apps w/text to voice), and move to streamlit where caching is super easy (has also improved in shiny a couple years ago but harder). Just thinking out loud in case someone else wants to make a public dashboard using the BigQuery data (or in case @pealco or others have a better way about it) 🙂