As a designer at Obata Noblin Office, I developed physical models as analytical tools — helping teams test massing, program, and spatial strategies — and managed the model shop through all phases of production. Working in cast concrete, walnut, beeswax, and CNC-milled foam, the models ranged from 1:200 massing studies to 1:20 sections.

Across projects I contributed to early concept work as well as later design stages, including spatial coordination and technical development, often through the lens of physical model making.

For Seal House, a California coast project, a late-stage change in beam depth required a clear spatial study: a 1:20 sectional model with a cast concrete base, and wood and paper articulating the structure and interior volumes. For Carriage House, designed for an architect in Idaho, a sectional model combined poured concrete, wood, and layered paper to communicate construction and material logic. For Willet, a coastal residence shaped by flood resilience and its sensitive site, I tested approaches from paper and cardboard to a beeswax cast of the terrain using a CNC-milled foam mould.

This studio, titled "Small Language Model," challenged us to rethink architectural design and representation by working as if we were an algorithm trained on only two precedents. Every design decision had to be informed solely by these precedents, as if they were our entire architectural knowledge.

The goal was to design a five-unit residential building in Tokyo, drawing inspiration from House in Midorigaoka (Itsuko Hasegawa, 1975) and S-House (SANAA, 1997).

The final scheme contrasts an open ground floor with a compact, grid-aligned upper floor. A slight rotation in the grid generates secondary spaces that support circulation and services. This rotated grid influences both plan and section: light enters through a translucent polycarbonate facade on the second floor and a central skylight penetrates to the ground floor. Two separate entrances provide private access to each unit, while staircases run between the interior grid and exterior envelope. Upper-floor windows extend beyond the grid to draw in daylight. Material choices — precast concrete and polycarbonate — reference the precedents and underscore the project's balance between solidity and lightness.

Werk Null marks the starting point of Holcim's cement production process and introduces visitors to the material origins of the industry. Located on the former rubber storage site in Siggenthal, the building functions as a workshop, exhibition space, and public interface between the factory and its surroundings.

The east and north façades are formed by a slender concrete frame infilled with steel gabions containing limestone sourced directly from the nearby quarry. This assembly makes the raw material of cement physically present while referencing the structural language of early industrial concrete architecture.

Inside, a single hall unfolds beneath a curved roof that echoes the factory's silos, unifying workshop, exhibition, and gathering spaces into a continuous spatial sequence. As both building and exhibit, Werk Null translates industrial production into an immediate, bodily experience and establishes a clear, public threshold to the site.

With Jonas Baltensberger and Mathis Steinmann.

In the summer of 2023, I received the Judith Lee Stronach Travel Scholarship, awarded to a small group of students to support independent research and service work. I used the opportunity to explore the architecture of refugee camps, a subject largely absent from architectural education despite its global relevance. My research focused on how architecture functions as both a tool for survival and a means of control, as spatial arrangements often limit movement, enforce dependency, and regulate daily life.

The project began at the Venice Architecture Biennale, where Foster + Partners exhibited deployable shelters made from fiber-reinforced cement fabric. I then traveled to Israel and the occupied West Bank, where I was welcomed by a community organization in Hebron in exchange for teaching English. These connections led me to Al Fawwar Camp, established in 1949 and now home to over 12,400 people, with a population density more than four times that of New York City. Over time, its temporary shelters evolved into dense, permanent communities shaped by necessity and restriction.

Upon returning, I exhibited my work at the UC Berkeley College of Environmental Design Gallery, combining photographs with research from humanitarian and academic sources to frame architecture as both a response to crisis and a mechanism of control. This became one of the most formative experiences of my education.

In 2020, I began designing a sleeping structure after purchasing 10 metres of fabric scraps from a Swedish tent company, prompting a six-month process to build a tunnel tent. The design consists of two parallel arches with fabric tensioned between them. The inner tent includes a waterproof floor and ultralight nylon panels that act as wind and visual barriers, while a transparent insect mesh at the top allows for ventilation and visibility. Elastic loops hold the poles away from the inner tent, and the outer tarp sits directly on the poles to provide weather protection.

Following the tent, I developed a more adaptable sleeping system: a modular platform for the rear of a Honda Element. After full-scale cardboard prototypes, I finalised a CNC-cut plywood design modelled in Fusion 360, which I taught myself for both CAD and CAM workflows.

The platform comprises three interlocking base panels with cross-lap joints, forming a rigid but lightweight frame. A central slot allows skis to pass through, while circular voids reduce weight. Two hinged panels fold out to form a flat sleeping surface and collapse for compact storage. The result is a stowable, vehicle-integrated system used extensively for travel across the western U.S. and Mexico.