For your scenario you need to have unique identifiers or update_at columns in your source system tables. If this is the case you can do incremental loads. Easiest way is to lookup the highest id or updated_at using a max function from your bronze layer. Next, you start reading the data from the source using the returned value from your bronze layer. An alternative approach would be using your metadata control table for keeping track of what records are processed.
When using the identifiers you must ensure that no older records are updated within the source. If so, these won’t be fetched.
An alternative solution would be to use CDC or alternative tooling that captures the delta for you. Hope this addresses your concern.