I also used Pachyderm pre-MLOps (2017+), specifically for time series ML (finance) and didn't see a mismatch between what we wanted and how Pachyderm worked. We

I also used Pachyderm pre-MLOps (2017+), specifically for time series ML (finance) and didn't see a mismatch between what we wanted and how Pachyderm worked. Were you overwriting/replacing data over time?