I think one problem that arises from practical work is: Databases seem to be biased towards either transactional (including fetching single records) or aggregat

I think one problem that arises from practical work is: Databases seem to be biased towards either transactional (including fetching single records) or aggregational workload, but in reality both are used extensively. This also brings difficulty in data modelling, when we DEs are mostly thinking about aggregating data while our users also want to investigate single records. Actually, now that I think about it, we should have two products for the users, one let them to query single records as fast as possible without hitting the production OLTP transactional database, even from really big data (find one record from PB level data in seconds), one to power the dashboards that ONLY show aggregation. Is Lakehouse a solution? I have never used it.