The idea behind CarbonDB is that a lot of components in your stack produce different amounts of CO2 based on how much energy they consume, how long the hardware is used, what time and where certain operations are computed and many more. For companies it is vital to record all this data so they can a) improve their stack and also b) report the data as will be required by new EU laws.
The solution is a central database in which all the different services can report their usage and then there is a central point where analytics and further operations can be performed
You can read all about it under: https://www.green-coding.io/projects/carbondb/