Making Omi a True AI Companion for Consumers

• 

  Product Mockups (from left to right: OMI SCORE, Home Page, Marketplace UI)

• 

  Product Mockups (from left to right: Memories page, Conversations page, Tasks page)

Inspiration

Omi is a powerful product, but during my conversations with the founder, one thing became clear: For consumers, this app without guidance feels overwhelming.

The app today does a lot extremely well: recording, transcription, task extraction, scoring, mentoring. But these capabilities were fragmented across the experience, making Omi feel more like a technical system than a daily companion.

This project started with a single question: How might Omi feel intuitive, guided, and habit-forming for everyday consumers..... without losing its technical depth?

The Track Challenges my product addresses:

1. https://github.com/BasedHardware/omi/pull/3790
2. https://github.com/BasedHardware/omi/issues/3121
3. https://github.com/BasedHardware/omi/issues/3577

What it does

This project reframes Omi’s core experience around clarity, execution, and guidance by:

1. Reducing onboarding to 3 high-intent steps to minimize drop-off
2. Reframing the home experience to clearly communicate system state and next actions
3. Elevating OMI Score from a secondary feature into a core, visible driver of behavior
4. Aligning tasks, conversations, and scoring into one coherent execution loop

The goal was to make existing strengths feel obvious and useful (but for non-technical consumers as well)

How I built it

This was a design-led sprint, focused on product thinking rather than production output.

My approach:

1. Founder conversations to understand 2-year product vision
2. Heuristic analysis focused on cognitive load and consumer usability
3. Rapid iteration on onboarding, home, and tasks experiences
4. Focusing on behavioral framing
5. Design decisions grounded in habit-formation and execution psychology

Challenges I ran into

1. Balancing technical power vs consumer simplicity
2. Avoiding feature bloat while still respecting Omi’s depth
3. Making OMI Score feel motivating
4. Creating guidance without making the product feel restrictive

Accomplishments that I'm proud of

1. Reduced onboarding friction while increasing clarity
2. Successfully repositioned OMI Score as a core product pillar
3. Improved visibility into tasks, execution, and system state
4. Created a more consumer-friendly mental model of Omi
5. Aligned closely with founder vision rather than personal preferences

Most importantly, my work did not redesign Omi... it simply clarified it.

What I learned

1. Execution is the clearest behavior-change promoter
2. Guidance builds trust faster than flexibility
3. Small structural changes can unlock large perception shifts

What’s next for us

If this were taken further, the next steps would be:

1. Validate design decisions with real user behavior
2. Measure task completion rates and daily engagement
3. Also focus on Mentor Mode to a personalized, adaptive layer with OMI Score
4. Explore habit-forming nudges tied to OMI Score trends

Built With

• figma