An investigation story: a client's AI image feature failed for most of every day, on the same daily schedule, since the day it launched — a Gemini per-day quota that drained early each day. Because it had always behaved that way, the team called it normal and quietly lost the users who hit it. How an ApexData audit found the pattern, and why a defect present from the first deploy is the hardest kind to see.
How to predict what PostgreSQL would do with a query without running it — the statistics the planner reads, where they live, and how an offline analyzer reconstructs the plan from collected pg_class, pg_indexes, and pg_stats data, plus the honest boundary between structural prediction and the live cost model.
Notes from a Tel Aviv meetup talk on AI-first coding in 2026: why developer skepticism is mostly outdated, where the real concerns sit (trust, security, fun), and how to share adoption inside a team without mandating it.
Claude Code Workshop & Best Practices Speaker: Evgeny Potapov, ApexData co-founder & CEO
Once the layers from Part 1 are in place and the code from Part 2 has been written, the on-call experience changes. Part 3 of a series on observability strategy: the practices that reduce the rate of escalations from on-call rotations to the developers who built the system.
Observability is not a tool you buy; it is code you write, and a meaningful fraction of the code in a working production system. Part 2 of a series on observability strategy: how much code is instrumentation, what that means for how a team works, and what it costs in development time.
An observability strategy designed from the people the system serves, not the boxes it runs on. A top-down tour of the layers — user experience, business signals, tracing, service monitoring, infrastructure, user feedback — and what each one answers.
