Baking Assistant

My Role: Founder/Lead Product Designer

Project Type: Mobile App Design with ML Integration

Focus: Computer Vision & Material Analysis

Outcome: High Fidelity Prototype

Back to Projects
View All Work
Explore More
Case Studies
  • Problem
    Find and interesting use of TensorFlows capabilities.
    Solution
    Create a fun mobile appthat figures out what kind of cake to put in your vessel.

  • Discovery

    Diving into the capability of TensorFlow and exploring its potential applications.

    In Situ
    • TensorFlow API Documentation: Looking through the TensorFlow API documentation to understand its capabilities and limitations.
    • Brooklyn Pottery: Working with a local studio to explore practical applications of TensorFlow in a real-world setting.

  • Competitor Analysis

    Analyzed competitors to identify industry standards and best practices.

    • Pinterest: Evaluated visual discovery and recipe sharing features for inspiration curation.
    • Tasty: Evaluated step-by-step video guidance and ingredient-based recommendations.
    • Yummly: Evaluated personalized recipe recommendations and dietary filtering capabilities.
    • Whisk: Evaluated recipe scanning and ingredient sourcing integration features.
    • Google Lens: Evaluated image recognition and search capabilities for visual discovery.
    • IKEA Place: Evaluated augmented reality features for visualizing furniture in real-world spaces.
    • YouCam Makeup: Evaluated personalized makeup recommendations and virtual try-on capabilities.
    • PlantSnap: Evaluated plant identification and care recommendations through image recognition.

  • Simple Task Flow

    Documented the process for early issue identification.

    • Scan object: Find Objects in the real world through sensors.
    • Find objects in Database: Using Tensor Flow find objects and match them with database entries.
    • Output Recipe: Output a recipe based on the object and its properties.

  • Quick Wireframes

    To get moving quickly I created wireframes that detailed out my ideas.

    In Situ
    • Clear Camera View: Ensure proper lighting and angle for accurate object scanning.
    • Simple Recognition: Display recognized object and confidence level to user before proceeding.

  • Object Recognition Flow

    After understanding the TensorFlow model, I designed a flow that accurately identifies objects..

  • MVP ICPACIT Flow

    Implemented simple, informative cards to display pipeline health, enhancing efficiency and accessibility.