Hi 👋🏽 I'm Vasu.
UX design leader for complex, high-trust systems.
I shape how 200M+ customers experience delivery on Amazon, and I design for the moment AI meets human judgment: from the signals a model reads to the moment it hands a decision back to a person.
High stakes. Real constraints. Shipped.
I design at the constraint layer — where business rules, user needs, and system limits collide. Every decision is an experiment. Every result is information. The craft is in knowing what to do with both.
Vibe Coding. Cautionary laws.
I. The Comprehension Trap
If the AI writes it faster than you can read it, you now own code you don't understand.
II. The Drift Law
Every prompt builds on the last. By prompt seven, you're solving a different problem.
III. The Debug Paradox
The more you ask the AI to fix its own bug, the more bugs you'll need to fix.
IV. The Translation Gap
What you said, what you meant, and what it built are three different things.
V. The Confidence Illusion
It will sound certain. It will be wrong. These two facts are unrelated.
VI. The Ship Trap
"It looks right" is not a test. It is a prayer.
VII. The Scope Creep Corollary
The feature you didn't ask for will break the one you did.