Hi @Paweł Haubus,
I've reached out to the product team for additional insights. Accordingly, there are several solutions which are dictated by the boundaries of your workstream.
- If you just removing sensitive data, you can create an ingestion transformation rule that remove the fields/values.
- If you want to allow access to some users and not others, you can split the incoming data into two workspaces. Similar to what you've done thus far but you don't need to create two separate subscriptions to achieve this. You would essentially filter the data on one and leave as-is on the other.
- If the non-sensitive data consumers are looking for less granular data, you can use a summary rule to create clean and aggregated summaries of the original data assign different RBAC roles to uses for the results.