Add persistent processing job records

[Dadud]

Summary

This PR starts Phase 3 of the modernization roadmap by adding durable processing job records underneath the current in-memory transcription flow:

• adds a call_jobs migration with status, attempts, priority, payload, result, and timing columns
• adds a src/jobs/processingJobs.js helper for creating and updating job records
• creates a transcription job when a transcription record is inserted
• marks transcription jobs as processing, completed, or failed around the existing storage/transcription path
• adds tests for migration planning and job helper serialization/status constants

Why

Scanner Map currently relies heavily on in-memory processing state. This PR does not replace that queue yet; it adds durable job visibility first so future PRs can safely move transcription, address extraction, geocoding, and publishing onto resumable job processing.

Verification

• npm test
• npm run check:syntax
• npm run check:config

Notes

This is stacked on top of PR #11, which is stacked on PR #10 and PR #9.