I can say that even now lots of machine learning datasets are kept as rows in gSheets and processed by Python running in Colab. In our team, all our evaluations: read input data from gSheet, run model inference, store outputs in gSheet for manual review if needed, process outputs to compute metrics and plot graphs in Colab. It is great because evaluation logic frequently changes, and without Colab you would sit and wait days each time for code reviews and stupid readability approvals.
And then, of course, if you want reproducibility, you just check in Colab notebooks into the source control.
I have heard that at google collaboration happens with:
- making lists (google groups)
- gsheets
- basic real time chat with auto disappear
Is that the case? I find that super interesting. No sexy Slack or Teams type of thing?
Chat supports rooms with chat threads without autodisappear. Everyone is in various rooms for their local team, broader team, org, and cross functional stuff. The net effect is Slack-like.
There's also the ticket system. Sheets are commonly used by various PMs and TPMs for high level tracking, but IC eng stick with the tickets.
It works fine.
This was not because it was a Google product (we used plenty of competitors' products) but because it is so easy to make them good enough for the task that you can move on to getting the job done instead of administrating getting the job done.