Blog
Field notes from shipped engagements.
Written by the engineers, designers, and PMs who shipped the work — not a content marketing agency.
AI
2026-04-22 · 8 min
How we run an eval suite for production agents.
Demos pass on friendly inputs. Production breaks on adversarial ones. Here's the eval harness we wire into every agent — and what we look for in the failure modes.
Web
2026-04-08 · 6 min
WordPress to Next.js: the redirect map we review page-by-page.
Replatforming without losing SEO equity is mostly a redirect map problem. Here's the workflow we use, the categories we check first, and the post-launch monitoring that catches the misses.
Performance
2026-03-19 · 5 min
Performance budgets in CI — fail the PR, not the launch.
We set the LCP/INP/CLS budget at sign-off and enforce it on every PR. This is the GitHub Action we use, the thresholds we picked, and what the team agreed when we set them.
Software
2026-03-04 · 10 min
Row-level security: the multi-tenant pattern we ship every time.
Tenant isolation in Postgres RLS, not application code. The trade-offs, the gotchas, and the audit story it lets you tell. Plus the migration we run when we inherit a multi-tenant codebase that puts isolation in the app layer.
Process
2026-02-18 · 4 min
Two-week sprints — the cadence at which surprises stay small.
Why two weeks, what a sprint demo looks like, and the retro template we use the Monday after. Not dogma — just the pace at which surprises stay debuggable.
Want one in your inbox
Subscribe to field notes.
One post a fortnight. Written by practitioners. We don’t send anything else.
Or grab the RSS feed.