We are currently working together MongoDB for digital shadow, where I think MondoDB really excels in. But for time-series data we o sously prefer timescaledb. I think performance does not count if it is just 20-30 percent better than any other tool (influxdb vs timescaledb, sorry James), but according to the article there seem to be 10x or even 1000x factors difference, which is significant. Please also note that the article is from TimescaleDB, who cannot be entirely neutral, but it would fit my general view on timescaledb and mongodb. In summary, I'd choose timescaledb for time-series data and mongodb for digital shadow (referencing then the timescaledb)