Skip to content

Backend Starter

If I’m hiring a backend developer and I want them to be immediately productive, there are a few non-CS fundamentals I would require expertise in, depending on the maturity and scale of the system. Here are the most impactful backend-specific domains that I’d lean heavily on:


🚨 Must-Have Domain Expertise (Beyond CS Basics)

Section titled “🚨 Must-Have Domain Expertise (Beyond CS Basics)”

1. Data Modeling + Query Optimization (SQL)

Section titled “1. Data Modeling + Query Optimization (SQL)”
  • Not just CRUD.

  • Must be able to:

    • Design normalized + denormalized schemas based on use case.
    • Understand query plans and indexes.
    • Optimize slow queries.
  • Why: Backend work is often bottlenecked by bad data access patterns more than bad code.

  • Knows REST vs GraphQL tradeoffs.

  • Understands:

    • Contract stability.
    • Pagination, filtering, error handling.
    • Breaking vs non-breaking changes.
  • Why: Poor API design causes years of tech debt and pain.

3. Authentication + Authorization (OAuth2/OIDC/Role-based access control)

Section titled “3. Authentication + Authorization (OAuth2/OIDC/Role-based access control)”
  • Knows how token-based systems work (JWT vs opaque).
  • Can implement scopes, roles, tenants, etc.
  • Why: Backend devs deal with access control daily, and mistakes here lead to security breaches.
  • Fluent in logs, traces, and metrics.
  • Can instrument code with structured logs.
  • Why: Debugging distributed systems is brutal without this. It’s where junior devs waste the most time.
  • Knows when to offload to background jobs.
  • Has experience with queues (e.g., RabbitMQ, Kafka, SQS) and workers.
  • Why: Backends need to scale beyond just HTTP request lifecycles.

  • Knows what to cache (and what not to).
  • Can use Redis, CDNs, in-memory caches wisely.
  • Understands invalidation techniques.
  • Knows CAP theorem, consistency models.
  • Familiar with leader election, retries, idempotency, etc.
  • Especially if the system spans multiple services or regions.

- Containerization + Deployment (Docker/K8s)

Section titled “- Containerization + Deployment (Docker/K8s)”
  • Can write a clean Dockerfile and debug container failures.
  • Doesn’t need to be infra-heavy, but must understand how their app runs in prod.

Summary: If I had to pick one, I’d require data modeling + query optimization. It has the highest ROI for backend dev productivity, especially in systems that touch any non-trivial data.