Supporting documents

Research and framing

Model, AI rules and content

The voice guide has its own documentation: Northstar Voice Agent | Conversation design

<aside> 🧭

Northstar is an interactive learning tool. Students build an imaginary AI, run it, and watch who it leaves behind. A narrator and a real-time voice guide explain the numbers in plain language. I took it from brief to a live web app on my own.

</aside>

Role Sole product designer: research, IA, interaction, visual, content and voice
Type MXDAI project, live product
Methods AI Experience Canvas, persona, jobs to be done, conversation design
Deliverables Live web app, IA and flows, content for 16 domains, voice guide, case study
Tools Figma, FigJam, Lovable, Claude, Deepgram
Links Live app · Case study · FigJam information architecture

1. Context

AI now helps decide who gets hired, who gets a loan and who gets seen first at a hospital. Most people learn about bias in these systems from articles and lectures. AI systems are often treated as neutral, while biased data and design choices quietly shape outcomes. The resources that explain this are either too technical or too abstract to connect with real decisions.

2. Problem

Tool Built for What it asks of you Gap for learners
IBM AI Fairness 360 Data scientists Python, fairness metrics, your own model Needs code and statistics to start
Google What-If Tool ML practitioners A trained model and a dataset Powerful, but assumes you know what to look for
Articles and courses Everyone Reading Explains bias without letting you cause it
Northstar Students and curious people Moving a slider and pressing Run Designed to close this gap

3. Design challenge

How can someone with no technical background cause bias, see who it affects and understand why, in a few minutes?

Who it's for. Primary: design and CS students, early-career professionals, and high-school students in AI or digital citizenship programs. Secondary: educators, workshop facilitators and high-school teachers adding AI literacy to their classes. Used in classrooms, workshops, online modules and self-study, often projected and discussed in a group.

4. Design principles

  1. Make the invisible visible. If a learner changes something, the impact shows immediately.
  2. Explain, don't lecture. Short, human messages that answer what changed and why it matters.
  3. Learn by doing. Try this, see the outcome, understand the implication.
  4. Design for the beginner mind. Approachable, low-jargon and safe to experiment with.
  5. Connect data to human impact. Show people, not only numbers.