Engineering

Backend

Working notes on backend engineering — Python services with FastAPI and Django, Postgres at scale, Redis, durable workflows with Temporal, event-driven and microservice architecture, and patterns for building reliable AI workflows and integrations.

11 subtopics

Subtopics11

01codePython Service FundamentalsAsync vs sync, type hints, project structure, and config — the foundations every Python service rests on.1 item

02zapFastAPIAsync-first APIs: dependency injection, Pydantic validation, async correctness, and production serving.1 item

03layoutDjangoBatteries-included CRUD: where Django wins, ORM discipline and the N+1 trap, safe migrations, async limits.1 item

04databasePostgreSQL — CoreThe day-to-day database craft: indexing, reading EXPLAIN, transactions and isolation, locking, data modeling.1 item

05trending-upPostgreSQL — ScalabilityScaling one Postgres, in order: connection pooling, read replicas, partitioning, vacuum tuning, then sharding.1 item

06layersRedisThe in-memory multitool: caching, rate limiting, distributed locks, queues and streams, ephemeral state.1 item

07repeatTemporal — Durable WorkflowsDurable execution for long-running and AI workflows: workflows vs activities, signals, retries, watch-outs.1 item

08git-branchEvent-Driven ArchitectureDecoupling with messaging: broker choices, the transactional outbox, idempotent consumers, ordering, DLQs.1 item

09boxesMicroservicesMonolith-first thinking: drawing boundaries, data ownership, sync vs async comms, sagas, resilience patterns.1 item

10sparklesAI Workflows & IntegrationEngineering around LLMs: reliability and retries, cost control, structured output, orchestration, RAG, observability.1 item

11gaugeScalability PatternsCross-cutting scaling: statelessness, caching layers, read/write splitting, back-pressure, autoscaling, the order to apply them.1 item