Why connection pooling is non-negotiable for Postgres apps. PgBouncer transaction-mode quirks, modern alternatives (pgcat, Supavisor, Hyperdrive), and the patterns from production.
Why gRPC over standard load balancers misbehaves, the right balancing modes (client-side, mesh-aware), and the patterns from production gRPC fleets.
When gRPC actually wins for service-to-service traffic. Performance numbers, the ergonomics gap, debugging tradeoffs, and how to mix gRPC inside with REST at the edge.
Why most health checks lie, the difference between liveness and readiness, dependency-aware checks, startup probes for slow boots, and the patterns that surface real problems.
Why most products should start with a modular monolith, when microservices earn their cost, the modular monolith pattern, and the modern split decision.
What changed in Go 1.24 for production backend code. Generics maturity, range-over-func iterators, swiss-table maps, weak pointers, slog improvements, and the patterns to use them.
How to use gRPC’s three streaming modes correctly. Server-streaming, client-streaming, bidirectional, deadlines, errors, flow control, and concrete patterns for live updates and uploads.
Picking an API versioning strategy in 2026 (URL vs header vs date), deprecation timelines, sunset headers, and the patterns that survive long-term customer relationships.
How to design webhooks producers AND consumers can rely on. Signing, retries, idempotency, ordering, observability, and the patterns from real production webhook systems.
Three multi-tenant patterns: shared (tenant_id everywhere), pooled (schema-per-tenant), per-tenant DB. The decision matrix by scale, isolation, compliance, and cost.