Start with the decision.
Before choosing a solution, agree on the user problem, business intent and constraints. A feature request is an input, not the whole problem.
What must we decide before building?
WORKING PRINCIPLES
How I make the work clearer.
The questions I use to connect product decisions, engineering work and release confidence.
Write behaviour that can be tested.
Use acceptance criteria to expose edge cases: retries, time boundaries, partial failures and access rules. Keep the product rule alongside the work item.
How would we know this works?
Automate with review points.
AI can support analysis, internal tools and everyday workflows. I connect system building with practical use and training, while keeping product intent, priorities and human review visible.
Which decision still belongs to a person?
Show the limits of verification.
Separate an automated pass from real-device checks, business acceptance and production observation. Record open risk and ownership before release.
What has not been verified yet?
SECURE PRODUCT THINKING
Trust belongs in the product conversation.
Access, data integrity and availability affect the experience. I bring these questions into product thinking without treating them as a substitute for security expertise.
A useful next conversation