Short technical notes on reliable AI-enabled software engineering.https://markhingston.dev/en-gbhttps://markhingston.dev/notes/deterministic-where-possible/https://markhingston.dev/notes/deterministic-where-possible/A useful boundary for AI systems is to keep judgement in the model and move invariants, validation, and repeatable transformations into code.Sat, 05 Sep 2026 00:00:00 GMTreliabilityagentic-systemshttps://markhingston.dev/notes/evaluation-is-system-design/https://markhingston.dev/notes/evaluation-is-system-design/Evaluation becomes more useful when it is designed with the workflow, not added after the model already appears to work.Fri, 04 Sep 2026 00:00:00 GMTevaluationllm-systemshttps://markhingston.dev/notes/convergence-over-capability/https://markhingston.dev/notes/convergence-over-capability/The next framework, agent, or tool is only useful when it resolves a concrete blockage better than the mechanism already in hand.Thu, 03 Sep 2026 00:00:00 GMTsoftware-designdeveloper-tooling