ChefMate

Perserving Family Traditions with Technology

Role

Design Lead

Project Manager

TOOLS

Figma

Miro

MockFlow

TEAM

Design Lead

UX Lead

Research Lead

User Testing Lead

TIMELINE

Oct 2024 - Dec 2024

(3 Months)

Description

Preserving what's lost between generations

ChefMate is an AI-powered mobile application that transforms cooking demonstrations into detailed, structured recipes by analyzing video and voice recordings. It preserves family culinary traditions while making them accessible to younger family members through modern technology.

Traditions are the threads that weave generations together.

Traditions embody shared history, values, and connection. Yet as families evolve, these traditions risk disappearing entirely, taking with them the stories, knowledge, and connections that define family identity.

We set out to answer a critical question:

"How might families overcome generational barriers (language differences, cultural shifts, and technological divides) to preserve meaningful traditions and maintain a sense of continuity across generations?"

My Role

Design Direction & Project Coordination.

ChefMate was developed as a group project for INFO 360: Design Methods at the University of Washington. Although my involvement was constant, I guided the overall design direction, facilitated design critiques, and ensured we met project milestones while maintaining a cohesive vision for the product.

Research

Understanding needs across generations.

Our research aimed to understand how different generations perceive and preserve family traditions. We conducted interviews with 8 participants:

  • 4 older adults (ages 65+) from diverse cultural backgrounds

  • 4 younger adults (ages 18-28) with varying relationships to their cultural heritage

Here are some of the types of questions we prepared:

Synthesizing research into clear patterns.

After grouping similar observations into meaningful patterns. Three clear themes emerged:

Cooking emerges as the universal family language.

A consistent pattern emerged and stood out across all participants: cooking is a universal tradition. This wasn't limited to recipes alone but included:

  • Family stories shared during cooking sessions

  • Techniques shared through demonstration rather than words

  • Cultural context behind ingredient choices and preparation methods

When asked about how they learned family recipes, 7 out of 8 participants mentioned demonstration rather than written instructions. One participant's comment particularly resonated:

"My grandmother never measures anything. When I ask her how much of something to add, she just says 'until it looks right.' That's impossible to write down."

Personal validation through my own experiences.

Curious of our research, I asked my mother for her signature dumpling recipe, she immediately said, "Let me show you" instead of writing it down. Further supporting our idea of traditions relying on demonstration and unspoken knowledge.

These findings led us to a clear opportunity: What if technology could watch and learn cooking techniques just as a family member would?

This approach would address our core research findings by:

  • Eliminating technical barriers for older adults

  • Capturing unspoken techniques and subtle adjustments

  • Bridging language gaps through visual understanding

  • Preserving both the recipe and the stories shared while cooking

This observation formed the foundation for our AI-powered solution, allowing each generation to engage with traditions in the most natural way.

Ideate

Three key elements in preserving family traditions.

During our ideation session, we focused on this opportunity area and examined how recent advances in AI might address the specific challenges we identified:

  1. Communication Barriers: AI could observe and interpret unspoken techniques that older adults struggled to articulate, bridging language gaps through visual understanding.

  2. Documentation Challenges: AI could transform imprecise instructions like "a pinch of this" or "cook until it feels right" into structured recipes with more detailed guidance.

  3. Technology Divide: We could design an interface that would feel natural to both generations—video recording for younger users and voice interaction for older adults with limited tech experience.

The concept of ChefMate began to take shape. Rather than requiring older adults to adapt to technology, we envisioned technology that could adapt to their natural way of sharing knowledge—through demonstration rather than explanation.

Establishing core principles to guide our design.

To guide our approach, we established four key principles that directly addressed our research findings.

  • Invisibility: Technology should fade into the background during the cooking process.

  • Accessibility: The interface must be approachable, especially for older adults lacking digital literacy.

  • Preservation of Context: Capture the stories and cultural context that make family recipes meaningful.

  • Bridging Generations: Create shared experiences that connect older and younger family members through cooking.

ChefMate would act as a third party in the kitchen—watching, learning, and documenting family cooking traditions through video and voice, preserving recipes.

Design

Mockups with MockFlow.

Working closely with our UX lead, I created low-fidelity wireframes using MockFlow to visualize the core user journey. We kept these wireframes minimal to focus on information architecture and user flow.



The wireframes addressed our key challenges through specific design choices:

  • For Communication Barriers: We designed a video-first interface that emphasized demonstration over written instructions, with simple recording options prominently displayed

  • For Documentation Challenges: We structured the recipe view to include both precise measurements and space for contextual information about techniques

  • For Technology Divide: We designed a linear, step-by-step cooking flow with clear navigation and minimal cognitive load

Gathering initial user feedback to refine our approach.

We conducted evaluation sessions with the rest of our team and a few other peers in our class. These sessions helped us identify several important insights:

  • The basic flow made sense, but the entry points after recipe details needed clearer distinction

  • The step-by-step interface felt overwhelming with too many icons and buttons

  • The help functionality seemed disconnected from the main experience

To address the feedback, I analyzed our wireframes and annotated areas for improvement:



The annotations highlighted issues that needed to be resolved:

  • Clearer separation between steps through better visual hierarchy

  • Strategic placement of calls-to-action to guide users to next steps

  • Reduction of redundant elements that created cognitive overload

  • Better contextual placement of text in relation to images

Translating concepts to working prototype.

Using these insights, we developed a functional high-fidelity prototype that refined our approach:


User feedback was clear and humbling.

We presented our prototype to young and older adults to gain a wider perspective. They thought we were on the right track and suggested minor changes. However, two comments stood out.

"It looks like every other recipe app. What is different?"

"It feels like theres no life to it"

Developing distinct visual Identity.

We needed to develop a distinctive visual identity that would set ChefMate apart while enhancing usability.

We worked on creating a stronger, more cohesive design system.

The new design system gave ChefMate a distinctive identity and directly addressed our usability concerns by creating clearer visual hierarchy and more intuitive navigation paths.

Simplifying the experience.

Additional feedback from our professor led us to find that the recipe screens still contained multiple help icons and controls that could overwhelm users, particularly older adults.

Taking our AI assistant concept further, we consolidated multiple help functions into a single conversational interface that:

  • Significantly reduced visual complexity on the screen

  • Created a more natural interaction model similar to asking a family member for help

  • Provided contextual assistance that adapted to user needs

This approach simplified the experience while creating a supportive element that mimicked the guidance an older family member might provide while cooking together.

Testimonial

Panel of designers and peers

ChefMate was presented to a panel of 3 product designers specializing in technology for older adults and accessibility, alongside 2 professors and 65 other classmates. The feedback we received validated our approach and provided valuable insights for future development.

"This solution addresses the core tension between preserving tradition and embracing technology. The AI assistant concept creates a bridge between generations that feels natural rather than forced." — Senior Product Designer, Healthcare Technology

"ChefMate solves a real problem I've observed in my research with multi-generational families. The ability to capture tacit knowledge through video is particularly powerful." — HCI Associate Professor @ UW

Academic rcognition.

Our professor responded enthusiastically to ChefMate, sharing that it resonated with him on both a professional and personal level.

He encouraged us to continue developing the concept beyond the course. His recognition of both the academic merit and practical utility of our work validated the time and care we had put into the project.

Looking forward.

The encouraging reception of ChefMate has inspired us to consider how the concept could evolve. The panel emphasized the importance of human oversight in the process. Specifically, allowing users to edit AI-generated recipes before finalizing them to correct any misinterpretations. Future development will focus on enhancing this human-in-the-loop experience while implementing segmented recording to make the process more efficient for longer recipes.

Reflection

First time working from 0-1.

ChefMate was my first experience working on a 0-1 product design project, which presented both challenges and growth opportunities. Initially, I struggled with Figma's auto layout and prototyping, but as the project progressed, I became much more comfortable which allowed me to focus more on problem-solving rather than the technical implementation.

Moving forward, my most valuable takeaway was learning to separate my ego from my work. By removing myself from the equation, I turned criticism from something personal into a practical tool for improvement. This mental shift allowed me to objectively evaluate feedback and incorporate insights without defensiveness, leading to stronger design solutions. In future projects, I'll continue to view critique not as judgment but as a resource.