st23am/call-sync

To manage and execute integrations between Livevox and Voterfiles

CallSync

This web app handles:

• Fetching question and response ids, activist codes, and status codes to be used in our configuration system. This is done through the /configure/:candidate endpoint.

• Validating a particular configuration and ensuring that all questions to be responded to exist and have the indicated responses. This is done through the /validate/:candidate endpoint.

• Executing the configured sync on yesterday's calls, and writing the data to VAN or a CSV. This will include other systems as (this project)[https://github.com/justicedemocrats/osdi-proxy] advances.

Call States

When everything is going well, calls can be:

• unsynced. This is either because the script has not yet run for the day or the call occurred before we started syncing.

• in progress. This means the call has been read from Mongo, and is currently being configured to be posted to VAN (or another system).

• finished. This means the call has been successfully written to VAN or a CSV. The call's document will have a receipt of its sync, either a VAN canvass response identifier or the CSVs url on S3.

• ignored. This means the call is of a result type that we do not want to sync, most likely dropped calls.

• queued for csv. This means that the call is of a result type that we want to pay attention to but that we do not want to sync to the voter file due to VAN's pricing policy of 1¢/req above 1k reqs/day. These calls will be exported as a CSV if the candidate is configured to use the money-saving half-sync strategy.

If things went wrong, there are two places it could have happened:

• We could have sent invalid data to VAN
• We could have not been able to configure the post body correctly, likely due to misconfigured Livevox data.

This leads to the following values for sync_status: [unsynced (no attribute present), in_progress, finished, ignored, queued_for_csv, attempted_error, unattempted_error]

Running

This app requires a particularly configured Airtable sheet, MongoDB, API Access to Livevox, and an S3 bucket. If you're trying to get it set up, you should probably just talk to Ben!