Congratulations on the launch. I'll keep an eye on this. We used to build custom, turn-key machine learning products for enterprise. Recently, after playing wi

Congratulations on the launch. I'll keep an eye on this. We used to build custom, turn-key machine learning products for enterprise. Recently, after playing with things like Voilà, Streamlit, and Superset, we made it possible for our data scientists and ML people to show prototypes and applications right from the platform, without worrying about creating a VM, setting up the environment, scp stuff, create an application, configure a server, set up authentication, send a link to the client, etc. I can envision doing something similar with Evidence. Given it's markdown, could we imagine having a Jupyter notebook containing markdown cells that somehow use Evidence? Could this be a JupyterLab extension? I'm asking this because we have live collaboration / collaborative editing notebook on the platform, with access to external data sources such as S3 as if they were filesystems, so several people could collaborate on the same notebook, see cursors and selections of others, etc. Why not do that on Evidence work as well: - I start a notebook. Add a Markdown cell. Some magic, I can do whatever it is I can do to generate reports with Markdown. - Share the notebook with other users. We get together and work on that visualization/report. Tangent: Something that kind of sucks is that some clients send us a database dump as a file, plus all other miscellaneous data. We have to create a MySQL database from that dump. It's not a big deal, but we don't like it.