Application performance profiling, optimisation, and latency reduction.
Postgres EXPLAIN plans: node types, cost interpretation, and six patterns that kill query performance on large datasets.
Probabilistic drop rate limiting: uncoordinated enforcement bypassing Redis for 1M+ RPS with zero coordination overhead.
Java collections: computeIfAbsent, getOrDefault, removeIf, immutables, and Comparator chains that eliminate entire bug categories.
The features that changed Java: lambdas, streams, records, sealed classes, pattern matching, and virtual threads. Java 8 to 21.
How HTTP evolved from sequential text to multiplexed binary streams over QUIC. What each version solves and when to upgrade.
Production LLM API patterns: streaming, function calling, retries, token budgets, cost optimization, and observability for backend engineers.
B-tree internals, composite index ordering, GIN for full-text search, partial indexes, and preventing index bloat in production.
Java streams: pipeline internals, where streams beat loops, parallel stream traps, and when to use collectors vs teeing.
Java 21 virtual threads: M:N scheduling, pinning hazards, ThreadLocal pitfalls, JFR detection, and what migration really takes.
From 5-second Spring Boot cold starts to 50ms with GraalVM native images. The real gotchas, wins, and whether it's worth it.