Despite co-existing in one space, it is rarely possible for exchange students and local residents in Tallinn to form meaningful connections. This project investigates the structural conditions that produce this social siloing and proposes a relational service design intervention to address it.

It specifically looks at how newcomers get trapped in an isolated "Erasmus bubble," sticking close to other international students because breaking into local routines feels out of reach. By popping this bubble, the intervention builds an easy, natural bridge that connects both groups through their daily lives.

1. Triangulated Field Research: My discovery phase mapped systemic isolation by triangulating primary qualitative insights through multi-contextual autoethnography alongside deep-dive interviews with both international students and local civic stakeholders in Tallinn.

2. Speculative Co-Creation: I utilized co-creation workshops, behavioral field tests, and public space design probes to navigate psychological barriers, which helped me isolate the need for a material "Third Object" to trigger natural human connection.

3. Hybrid Strategic Framework: I used a hybrid design framework combining the Double Diamond and Stanford Design Thinking.

1. Different Rhythms of Life. Students arrive with the unburdened drive to explore, while locals are bound to pre-established professional and domestic routines.

2. Connection as a Byproduct. When connection is the primary goal, it fails to manifest. Instead, it appears as a byproduct of a shared engagement with an external purpose. For example, attending lectures or hobby circles.

3. The Third Object Concept. People can physically coexist in a shared space without ever interacting due to a lack of situational activators. Connection requires a shared tangible material or communicative catalyst that provides a reason for encounter.

Throughout the process, I learned how to apply Relational Design and Participatory Design methodologies to treat human relationships as my primary design material, co-creating the handbook's content and visuals directly with my research participants to transform them from passive users into co-authors of the system. Simultaneously, I applied the principles of Service Design and Social Infrastructuring to expand this physical artifact into a resilient, self-renewing backend ecosystem that shifts the focus away from isolated, temporary event planning toward a durable social grid.

This systemic approach helped me develop the capacity to build long-term, socio-technical frameworks that continuously coordinate people, spaces, and collective actions over time despite the constant turnover of student cohorts.

Wombet's AI had strong analytical capability but users were abandoning sessions early. Exit interviews revealed a consistent pattern: the interface made them feel like they were operating a machine, not having a conversation.

The brief was to redesign the interaction model so that casual fans could engage just as naturally as data-native power users — without dumbing anything down.

Mental model interviews with 18 users across three segments — casual fans, semi-professional analysts, and fantasy sports players. Each group brought entirely different frameworks for what "useful" meant.

The core insight: all three groups thought in narratives, not numbers. They wanted the AI to tell them *why* something mattered, not just *what* the data said.

The product's problem wasn't the AI — it was the conversation design. Responses were technically accurate but structurally opaque. We redesigned the response architecture around a simple principle: lead with the story, support with the data.

We also introduced progressive disclosure: high-level insight first, statistical support one tap away, full data on demand. This tripled engagement with the deeper layers without overwhelming new users.

AI UX is fundamentally conversation design. The quality of an AI product is determined not just by what the model knows, but by how that knowledge is shaped, paced, and delivered in context.

This project sharpened my understanding of how to design for different expertise levels simultaneously — and the importance of letting users set the tempo of their own experience.

Kuldīga faced a structural challenge familiar to small towns across Latvia: young people were leaving after secondary school and not returning. The municipality wanted to reverse this trend but had a limited budget and a historically top-down relationship with its youth population.

In addition, despite civic education being part of Latvia’s national curriculum, many young people remain disengaged from local civic life..

An initial stakeholder mapping exercise revealed that youth and municipal actors had fundamentally different diagnoses of the problem. The city thought young people left for jobs; young people said they left because nobody asked them to stay.

Interviews and surveys with 30+ high school students aged 16-18 helped to gather insights into their motivations, barriers, and perceptions of civic participation. Additionally, interviews with local employers, youth center, and teachers painted a far more nuanced picture of what "opportunity" meant to different groups.

1. Recognition & visibility: Students want to be seen and acknowledged by their peers and elders.

2. Access & convenience: Civic events must fit into students’ daily lives and be easy to find and join.

3. Gamified motivation: Tangible goals and point systems can boost intrinsic motivation.

From these insights, we reframed our challenge from simply “raising awareness” to creating visibility, peer relevance, and emotional reward around civic participation. This shift laid the foundation for a more youth-centered design strategy.

I strengthened my fluency in design tools and gained confidence in using service design frameworks like journey maps, blueprints, and community theory. Teamwork across cultures is powerful. Working with five other students from different countries expanded my perspective and sharpened my communication skills.

KuldiGO challenged me to fail fast and iterate thoughtfully. The pivot from our original idea to a functioning prototype only happened because we committed to testing early — even if it meant discarding work.