Zenstate

Process

The project followed an iterative design and development process where we continuously tested, evaluated and refined the concept. The team worked in parallel across design and development while regularly discussing decisions and improvements.

Idea & Concept

We started by exploring how data from wearable technology could be used in new types of mobile services. From this exploration, we identified an opportunity to combine health data with everyday planning. The initial concept was a calendar that could understand more than just what a user had scheduled. By connecting calendar events with information such as sleep, steps, calories and resting heart rate, the application could provide suggestions based on the user's current state.

Visual Design Choices

LoFi/HiFi-prototyping & User Testing

The design team created two separate lo-fi prototypes, which were tested using A/B testing and scenario-based testing with six participants. The testing helped us understand how users wanted to interact with their calendar and which information was most important. One clear insight was the need for a simple monthly overview and an easy way to add activities. Based on the feedback, we combined the strongest elements from both prototypes into a shared lo-fi concept.

HiFi-prototype

The combined concept was developed into a hi-fi prototype in Figma. We focused on creating a calm and accessible visual identity that would feel appropriate for users experiencing stress. The visual language uses soft colours, rounded shapes and the Jost typeface to create a simple and harmonious interface. We also reduced the navigation to three main areas to make the application easier to understand. The calendar was designed to support different levels of planning, allowing users to switch between month, week and day views. The hi-fi prototype was tested with four participants. Their feedback led to further adjustments to typography, contrast, navigation and the overall hierarchy of the interface.

Development

In parallel with the design process, the development team created the application structure, Firebase database and the foundation for the AI functionality. The application was developed using Flutter and connected to Firebase for authentication and data storage. Health data was accessed through HealthKit, while the device calendar was integrated through the device_calendar library. A key part of the development was the AI planning assistant. The assistant gathers information such as upcoming calendar events, health data and the user's request, and uses this context to generate personalized activity suggestions. The AI was implemented using the Groq inference platform and the openai/gpt-oss-120b model. The generated suggestions could then be converted into calendar events and added to the user's schedule. The final prototype was tested again using scenario-based tasks to evaluate the implemented experience and identify remaining issues.