A backend engineer's mental model for database indexing — what to index, the leftmost-prefix rule, and why a query still does a full scan.

How to design APIs and message consumers that survive retries safely — idempotency keys, HTTP semantics, and idempotent consumers.

How to scale a Django backend to 170K concurrent users — stateless services, autoscaling, load balancing, and protecting the database.

When to break a Django monolith into microservices — and how, with the strangler pattern, an API gateway, and a database per service.

Designing a workflow engine for processes that mix automation, human steps, and retries — using an explicit state machine, not tangled conditionals.

A practical framework for choosing search — start with PostgreSQL full-text, and move to Elasticsearch only when requirements demand it.

Building a resilient web-scraping pipeline: per-source spiders, hash-based incremental crawls, proxy rotation, and data-quality gates.

Building real-time fleet control with MQTT over mutual TLS — push not poll, per-device identity, and local-first enforcement.

Designing a policy engine where rules are data: authored and versioned centrally, evaluated locally so decisions stay instant and offline-safe.

How to model approval workflows that resolve dynamically from your org hierarchy — so a reorg never means a code change.