Aurum
Download

Backend Services Engineering

Server-side services built to be correct and easy to run: clear business rules, safe error handling and timeouts, background jobs and integrations that retry without harm, and the routines that keep them healthy.

Free · Mac (Apple silicon) and Windows · v1.0.0 · Other platforms

Then open the Marketplace in the app, search for Backend Services Engineering, and click Install. It goes into your own files, and works offline from then on.

What it installs.

1

Pillar

1

Area

50

Projects

201

Starter tasks

⚙️ Software Engineering

Backend Services Engineering

Building server-side services, business logic, background jobs and integrations that are correct, secure and easy to operate.

  1. Inventory of the backend services you own
  2. Trace of one request from edge to database
  3. Fail-fast configuration and secrets loading
  4. Error taxonomy and consistent failure responses
  5. Input validation at every service boundary
  6. Timeouts and deadlines on every outbound call
  7. Health checks, readiness and graceful shutdown
  8. Transaction boundaries and connection pool settings
  9. Domain glossary and written business rules
  10. Weekly dead-letter and failed job review
  11. Third-party provider changelog watch
  12. Scheduled job register and missed-run alerts
  13. Top exceptions triage each week
  14. Webhook delivery failure review
  15. Integration credential and certificate expiry calendar
  16. Monthly reconciliation against provider records
  17. Dead endpoint and stale flag sweep
  18. Backend support escalation queue
  19. Idempotency keys and safe retries in practice
  20. Race conditions and locking in business logic
  21. Transactional outbox for reliable events
  22. Money and currency handling in code
  23. Time zones, timestamps and calendar logic
  24. Domain modelling of invariants and aggregates
  25. Queue delivery semantics and message ordering
  26. Authorisation rules for business resources
  27. Choosing a background job framework
  28. Moving slow work out of the request path
  29. Retries, backoff and circuit breaking for a flaky dependency
  30. Outbound rate limits and quota budgeting
  31. Audit log of business-critical actions
  32. Build or buy for email, search or notifications
  33. Replacing provider polling with webhooks
  34. Direct-to-storage file uploads with signed URLs
  35. Personal data export and deletion handling
  36. Payment provider integration go-live
  37. Launch readiness for a new backend service
  38. Migration before a provider API version sunset
  39. Bulk data import for a new large customer
  40. Large notification send for an announcement
  41. First ninety days owning an inherited service
  42. Solo backend engineer at an early-stage startup
  43. Frontend engineer moving into backend work
  44. Freelance integration build for a client
  45. Taking over integrations from a departing colleague
  46. Backend work in a regulated domain
  47. Durable workflows for long-running processes
  48. Double-entry ledger for balances and credits
  49. Duplicate-safe event consumers with a processed log
  50. Tenant isolation in service logic

Where it starts.

Each project arrives with its purpose, milestones and first tasks written out. Here is how the first three read.

Inventory of the backend services you own

Most engineers can name the main service they work on but not the cron jobs, queue workers and small integration services that also page them. Listing every deployable, what it does, what it depends on and who answers for it turns vague ownership into a page you can plan from.

First tasks

  • List every deployable process from the deployment platform, not the repo list
  • Note the datastore, queues and external providers each one uses
  • Name an owner and backup owner beside every entry
  • Send the list of unowned processes to your lead with a proposal

Trace of one request from edge to database

Reading code file by file rarely shows how a request really moves: middleware that rewrites headers, an auth layer that loads the user twice, a hidden retry in the HTTP client. Following one important request end to end, with a debugger or a trace, gives you the mental model every later change depends on.

First tasks

  • Pick the endpoint whose failure would cost the business most
  • Step through one request locally with breakpoints at each layer
  • List each query and outbound call with its configured timeout
  • Raise a ticket for the most surprising thing you found

Fail-fast configuration and secrets loading

A service that starts with a missing payment key and only fails on the first real order is worse than one that refuses to boot. Loading every setting into one typed configuration object at startup, validating it and reading secrets from the secret store rather than files in the repository, catches mistakes at deploy time instead of in front of customers.

First tasks

  • Search the codebase for every place an environment variable is read
  • Move them into one typed configuration module validated at startup
  • Scan the repository history for committed secrets and rotate any found
  • Write the settings table with defaults and per-environment overrides

Make it yours.

Install Backend Services Engineering, then tell Aurum the date you're aiming for. It checks whether the plan fits the weeks you actually have.

Free · Mac (Apple silicon) and Windows · v1.0.0 · Other platforms

Blueprints are general information to help you organise your own work. They are not medical, financial, legal or other professional advice. See the Terms.