Thank you, is working great!
The idea is that i use a background job to aggregate some data (per month) and store that in DB.
To make the operation idempotent and because multiple jobs can run in the same time, i store an aggregated row per workerId. After that another job is summing all those aggregated rows per worker.