qventisqventis

Field notes for quality leaders

Release intelligence, packaged apps, architecture, agentic AI and the pitfalls that slow every test cycle. Filter by topic or search.

Reading a vendor release note like a tester

Vendor release notes are written for admins and buyers. How a tester reads them: arrival type, behavior verbs, configuration touchpoints and the gaps.

Oracle Fusion quarterly update readiness checklist

A practical checklist for each Oracle Fusion quarterly update, from reading What's New to signing off production, with the evidence to keep at each stage.

Reference architecture: release-aware QE

The parts of a release-aware quality practice: release intake, one app model, rule-based impact, deterministic engines, a signed quality view and CI hooks.

Flaky tests: classify failures by cause

Flaky is a symptom, not a diagnosis. How to classify every failed check by cause, from data and environment to timing and vendor change, using evidence.

Where requirements-to-test traceability breaks

A linked test is not a covered requirement. The five traceability gaps that hide in most matrices.

What agentic test automation should and should not do

Agents are good at drafting, exploring and triaging. They should not decide what a test means at run time or approve their own work. Where to draw the line.

Workday R1 and R2 regression without the weekend

Workday feature releases come with a preview window that many teams use too late. How to plan R1 and R2 regression so it ends on a weekday.

Human approval gates for AI-generated heals

AI can propose test heals quickly. Deciding which heals need which reviewer, what a proposal must show and how to keep the queue moving is the real design.

SAP S/4HANA upgrade testing pitfalls

Custom code that compiles but behaves differently, Fiori apps replaced by successors, interfaces tested only on stubs: the S/4HANA upgrade pitfalls we see.

Deterministic execution, AI only at authoring

Where AI sits in a test architecture decides whether results can be reproduced and audited. A design that uses AI to author and a compiled plan to run.