Thoughts on implementation of many separate multivariate time series?

[PorkShoulderHolder]

In my application, we have multivariate time series data (some signals at a consistent interval, others not) associated with each human user. What is the best configuration of these data in questdb if

1. Data from multiple users is rarely accessed at the same time
2. Application should scale as the number of users grow and as the number of samples for each user grows

I'm still learning about questdb and any ideas about this would be much appreciated