For more data stability we should split parsers into the layers: A Layer - pure chain data, just reflect transactions/events in local DB B Layer - the project-related contracts with data from contract variables/functions C Layer - calculated data based on B Layer