Improve Batch processing v2

[Jkd-eth]

User Story:
As a data engineer
I want to set up an internal batch scoring of addresses
so that I can process large datasets efficiently for the data team and provide results

Acceptance Criteria:
GIVEN the current system of batch processing
WHEN the bathc processing runs and there are errors
THEN there is an easy way to reprocess those erorrs quickly and efficiently and provide the consolidated results

This is a continuation of #3184

The batch process is working well and it's pretty quickly to pull results (400k in ~24-30hrs) however, it's an inconvenience for the Engineers & Data Scientists to have to manually pull & re-run the errors.

Update the DB connection to switch from dedicated connection to a data pool

Can we add the following once a batch process has run

• New UI in Django (see error messages, retry, re-run batch)
• Keep all records in the DB (store each address in DB)
• Automatically ‘re-run’ after the batch completion and only the error records are re-processed (e.g if a single network has failed on that network will be re-run
• Download the finished consolidated file with all wallets with a click of a button

Product & Design Links:

Tech Details:

Open Questions:

Notes/Assumptions: