Your prompt is what reads your mind.
Persona, job to be done, mental model, task and expectations give an AI teammate the context it needs.
A field note on designing with AI
A working notebook from Sapna Nayak’s session on using AI as a design partner—one useful piece of context at a time.
Persona, job to be done, mental model, task and expectations give an AI teammate the context it needs.
Don’t wait for one giant starting block. Add context, see the output, and let the work reveal what belongs next.
The useful shift is not “ask AI to design.” It is to bring AI into the room like a teammate. That means describing who is involved, what they need to do, what they already understand, and what a successful outcome looks like.
For this demo, the persona is our three-person team: a graphic designer, a UI researcher and a UX designer. The audience expands that circle to designers, PMs and general attendees who want to catch up and save the ideas for later.
The first edition stays read-only. Its information architecture is intentionally simple: these notes, shaped into a clear reading journey. Later, the same surface could connect to a database and become a living team memory.
For this event, the organiser’s post is part of the record. Read the organiser’s LinkedIn post

Persona, job to be done, information structure, mental model, task and expectations form a flexible prompting toolkit.


HTML can hold the canvas, document, idea, prototype and collaboration trail in one readable place.
