Node.js: fast, scalable JavaScript backend

Express is the most common but the loosest in structure; ideal for small projects to keep it lean. NestJS brings Angular-style modularity, great for long-term sustainability in enterprise. Fastify is performance-focused, preferred where throughput is critical.

We recommend NestJS in most cases because code structure stays consistent across a team — when a developer leaves, another can ramp up quickly.

Three ways to ensure type safety between frontend and backend: REST (manual type sync), GraphQL (auto-types via codegen), tRPC (TypeScript-native end-to-end types). For most Next.js + Node.js projects, tRPC is the fastest path.

REST or GraphQL fits when a public API is needed because tRPC is bound to the TypeScript ecosystem. The API consumer (internal tools or public API) drives the choice.

Node.js's event loop isn't suitable for long-running tasks. Sending emails, image resize, AI calls — all go to a background queue. We implement with BullMQ + Redis.

HTTP requests stay fast; background tasks run in parallel; failure handling (retry, exponential backoff, dead-letter queue) ships ready.

Monitoring Node.js performance in production isn't optional. Sentry for error tracking, pino for structured logging, OpenTelemetry for distributed tracing — all configured day one.

Bottlenecks are usually DB queries, external API calls, or wrong cache strategy. Caught proactively; alerts fire when needed.