Backend
Python and server-side interview topics, ordered as a learning progression. Start at the top if you are building coverage; jump straight to a phase if you are preparing for a specific role.
Where to start
If you have one evening, read Python core and Async and concurrency. They carry the most questions per page of anything here, and every Python interview draws on both.
The progression
Foundations
- Theory foundations — complexity, memory model, what the interpreter actually does
- Python core — the language itself: data model, typing, tricky questions
- Python OOP — classes, descriptors, MRO, dataclasses
Concurrency and testing
- Async and concurrency — asyncio, threads, processes, the GIL after free-threading
- Testing — pytest, fixtures, mocking, property-based testing
Web
- Web frameworks — FastAPI, Django, Flask, Pydantic
- REST APIs — REST design, versioning, pagination, idempotency
Data
- Databases — SQL, indexing, transactions, NoSQL
- Caching — Redis, invalidation, cache patterns
- Message queues — Celery, Kafka, RabbitMQ
Security and integration
- Authentication — OAuth2, OIDC, JWT, sessions
- Protocols — HTTP, gRPC, WebSockets, GraphQL
Design
- Architecture & design — layering, dependency direction, patterns worth naming
- Microservices Architecture — boundaries, sagas, service communication
- Observability — logging, metrics, tracing, OpenTelemetry
Infrastructure
- Docker — images, layers, multi-stage builds
- Kubernetes — workloads, networking, resource limits
- IaC — Terraform, declarative infrastructure
Cloud
- Cloud: AWS — the one most specs name
- Cloud: Azure
- Google Cloud
Tooling
- Git — the operations that come up under pressure
- Linux & bash — processes, signals, text wrangling
Practice and cross-cutting
- Data structures & algorithms — algorithms, in Python, at interview scope
- Security — OWASP, secrets, supply chain
- Code quality — linting, typing, review standards
- CI/CD — pipelines, deployment strategies
- Computer Networking — TCP, TLS, DNS, load balancing
Data platform
- Data engineering — Spark, pandas, Polars, Parquet, Delta, data architecture
Regulated domains
- Quant and fintech — market data, backtesting, portfolio optimization, FIX
- Healthcare and regulated industries — HL7/FHIR, PHI, GxP, clinical document intelligence
The two domain folders exist because job specs increasingly assume one. The engineering underneath is the same; the vocabulary and the failure modes are not.