No PM.
Undefined system.
10 days.

No PM.
Undefined system.
10 days.

I stabilized the architecture and shipped it at AFM.

I stabilized the architecture and shipped it at AFM.

Summary

White Rabbit Holes was a 10-day interactive system launched at AFM (American Film Market).


With no PM and no defined architecture, I defined the flow, state logic, and scoring gate to stabilize execution.

Product Type:

Interactive System

Context:

AFM Live Showcase

Role:

Product Designer

Architecture · System Logic · Scope Control

Constraint:

10 Days · No PM · Undefined States

Stabilizing the System Under Constraint

Launched live at AFM
0 demo failures
40% faster simplified flow
Clear scoring + allowlist logic defined
Team aligned within 3 days

Decision 01

Define the Architecture

The concept was cinematic.
The system had no defined sequence.


Without a clear sequence, engineering couldn’t build confidently, and the experience risked collapsing during live demo.


I gathered the team and mapped the full architecture before refining any visuals.

Flow architecture defined before UI refinement.

This turned a drifting idea into a buildable system.

Decision 02

Protect the Decision Gate

The most critical moment in the experience was not the teaser.

It was the scoring check.


Small human inputs determined:


– Access to reward
– Eligibility for allowlist
– Perceived legitimacy of the studio

I defined how the system:


– Evaluates correctness
– Assigns system state
– Triggers downstream logic

Human input becomes system state.

This prevented ambiguity and ensured consistency during live demo.

Human input becomes system state.

This prevented ambiguity and ensured consistency during live demo.

Decision 03

Reduce Risk Through Scope Control

With 10 days left, I cut non-core ideas to protect stability.


I removed:

  • Extra gamified layers

  • Decorative interaction loops

  • Late-stage feature ideas

I focused on:

  • Defined states

  • Stable scoring

  • Clear reward trigger


Scope control protected launch stability.

Prototype Snapshot

Design as Risk Management

Live systems don’t fail because they look bad.
They fail because logic is unclear.


I defined states and decision logic to protect execution.

Preventing System Drift

When direction keeps shifting,
nothing solidifies.


I fixed the structure first.
The rest followed.