Roofscape Haus draws its form from the mountainous landscape of Jecheon, Korea, and the region's heavy annual rainfall. The geometry of the roof is generated by the paths water takes as it sheds across the surface. The section follows the logic of the traditional Korean Hanok: a deep overhanging eave that blocks the high summer sun and welcomes low winter light deep into the interior.

A three-story structure, the building houses a design studio on its upper floors, with storage and a showroom for a timber manufacturing company below. Its primary frame is glue-laminated timber (glulam), which spans further than conventional wood and permits the complex sectional geometries the design requires. The exterior is clad in carbonized wood that retains the cellular structure of the timber while resisting cracking and weathering.

Most contemporary conversations about sustainability are framed around energy performance. I wanted Roofscape Haus to argue for a broader frame: one that holds engineering, environmental performance, and the regional and traditional architecture of its cultural context inside the same conversation. What does Korean cultural sustainability look like as a built proposition? That is the question this project is sitting inside.

Micro Housing II is a small, off-grid, prefabricated cabin designed for the seam between dense Korean urban life on weekdays and the open wilderness on weekends. The project draws on a Korean tradition that ties the mountains to ancestral ritual, and asks what a contemporary house designed to sit inside that relationship might be. It is a multi-family unit clad in zinc and wood with plywood interiors, small enough to feel modest in the landscape and complete enough to feel like home.

The interior holds a kitchenette, upper-level sleeping, a bathroom, storage, water heating, waste management, and power: everything an off-grid weekend household needs without growing beyond the cabin scale. The wall assembly pairs high and low thermal mass to keep temperature stable through seasonal extremes; the roof folds into the wall, treating insulation, glazing, shading, and ventilation as one continuous envelope rather than four separate systems. Exterior decks pull the everyday rituals of cooking, eating, and gathering out into the landscape, blurring the line between the architecture and the site.

Pedestrian bridge and urban deck designed as the central artery of the City Foundry redevelopment in St. Louis. The bridge sits on the Brickline Greenway, linking a mixed-use food court and shopping complex to the St. Louis Armory entertainment district. The bridge twists between two elevated interstate lanes (narrowing for vertical truck clearance eastbound, widening into LED-screen frontage for westbound motorists) and is clad in dichromatic and Fresnel lenses that split, collimate, and magnify the light passing through them, turning the structure into an optical instrument visible from I-64.

An engineered filter-gutter system collects polluted highway runoff and routes it into retention ponds planted with regional remediation gardens. At the City Foundry end, the bridge resolves into an amphitheater stage and a microclimatic canopy over a grove; the seating area doubles as farmers' market by day and performance venue by night. Below, the Urban Deck reads as an extension of the ground itself: a recreation complex with pool, basketball and racquetball courts, running tracks, gardens, food stalls, and the shared service amenities of a public commons (bathrooms, showers, playgrounds). After dark the upper deck becomes a dance hall and outdoor cinema, drawing density toward the new pedestrian routes.

The project is infrastructural in the broadest sense, holding engineering, landscape, architecture, and environmental science in a single civic gesture rather than divided across disciplines. An urban gateway is more than a sign of a city; it is a piece of its public life. The most useful thing a piece of mobility infrastructure can do in a fragmented downtown is make the ground walkable, the runoff drinkable, and the night worth showing up for.

Two facade experiments built around a single question: can a regular module produce irregular, atmospheric variation? Experiment 1 derives a precast brick from a smoothed knot (four round voids cut from a rectangular mass), yielding a module with six flat faces that can be stacked vertically, on edge, mirrored, or rotated. Experiment 2 begins from the venation pattern of a banana leaf, generating an interlocking curve that becomes a stackable, hollow-cored ceramic brick.

Both modules are tuned by a parametric gradient (frequency × mirror-plane rotation) that drives the wall from order toward randomness in measured steps. Experiment 1 is plaster cast in 3D-printed molds (Dremel 3D), with the mold system iterated until the seams fall out. Experiment 2 is printed wet on a Potterbot clay extruder; print parameters (nozzle, layer height, extrude speed, clay density) are tuned across seven prints to control slumping and cracking before firing. Each brick takes 21–32 minutes; a wall is hundreds of them.

Both experiments propose facades for dance pavilions. Experiment 1 lines a ceiling-to-floor mirrored passageway with sliding shading panels: soft, dappled south light against a continuous reflective wall. Experiment 2 becomes a 340-panel ceramic facade at East Beach, Santa Barbara, where the panels rotate on tracks so the building's skin is reorganized daily by its occupants. The argument is structural: randomness, made parametric and load-bearing, is not noise added on top of a discipline. It is the discipline.

A kid-scale geodesic structure built from 3D-printed connector nodes and stock dowels. Three node families (three-foot, five-foot, and six-foot) are derived from a single mesh that is remeshed for inner-surface curvature, weave-shelled into a thin solid, and cut to receive the rod ends. The same parametric system that generates a facade module produces a node geometry tuned to a specific span.

Each node is sized to a Doremi 3DA1 desktop FDM printer, with shells thin enough to keep print time under 36 minutes per node and material consumption to about 0.55 cubic inches. Print orientation, infill, and layer height are tuned so the node prints flat-down without supports, a constraint that keeps the kit reproducible on the smallest, cheapest printers a school or family is likely to own.

The argument is a quiet one about scope. Digital fabrication research often ends in a portfolio image; this project ends in a child's hands. The puzzle proposes that the most useful test of a parametric system is whether it produces something assemblable: a structure that holds itself up, fails in ways you can see, and survives being played with.

An urban design study of Shanghai read through Michel de Certeau's distinction between strategies (the top-down planning of corporations and the state) and tactics (the everyday, opportunistic uses of space that residents practice from below). The way in is hydrology: Shanghai's flood map traces the seam between Puxi's fine-grained historical fabric and Pudong's rapid corporate development, and reveals how indiscriminately the new ground absorbs rain.

The project locates the city's surviving tactical fabric in the longtang, the lanes and back alleys of central Shanghai that residents, vendors, and culture continually repurpose at small scale. In Pudong these lanes have been replaced by superblocks too coarse for foot, bike, or rain to move through. The diagnosis is not aesthetic: large, smooth, strategically planned ground is also the ground that floods, and it is the ground that strands its inhabitants when the weather turns.

The proposal is the Urban Blossom: a deployable canopy assembled from inputs by multiple stakeholders (architect, vendor, resident, choreographer) and reconfigured through the day and the year. The same structure becomes food market, weather shelter, performance ground, urban umbrella. The argument follows from de Certeau: the role of the urban designer is not to author the space, but to design the open frame inside which other actors can improvise. A tactic, not a strategy.

A reflective dialogue model that combines tactile tools, guided prompts, and AI-supported interpretation to help students clarify values, articulate intentions, and make informed academic decisions. The premise is that decision-making improves when students can externalize their thinking onto something they can move with their hands (physical cards, sortable prompts, sketches) before they are asked to commit to a sentence or a plan.

I'm building Dialogue by Design as part of the Career Design Lab at Dartmouth, alongside students and advisors who already meet weekly in workshop format. The AI's role is interpretive, not directive: it surfaces patterns across what the student has already said and made, and offers framings the student can accept, reject, or rewrite. In development.

Could a digital environment support the slower kind of thinking, not in a single sitting but across weeks? DartWorld is the long-running platform we built with the DALI Lab at Dartmouth to test that question. Most career counseling assumes students already know what they want and operates as one-off thirty-minute sessions; existing instruments (personality quizzes, job databases, alumni networks) treat identity as fixed rather than developmental. DartWorld starts from the opposite assumption: that what students want changes through exploration, and that a platform's job is to make that exploration durable across an entire term.

DartWorld is three connected components. A nine-question onboarding quiz that begins with origin and field of study and progressively turns inward, sorting students into one of three archetypes (Explorer, Seeker, or Achiever). A 3D navigable environment that lets students walk through occupations, industries, and stories from other students; career exploration as a place you move through rather than a list you scroll. And a journaling pinboard where students write, return to earlier entries, and watch their priorities shift over time. Development ran in iterative ten-week cycles with the DALI Lab, a student-led tech lab whose teams rotated each term. That constraint forced the system to be modular, with every decision written down for the next cohort.

Three findings shaped the build. The intake quiz turned out to be the richest source of data in the whole system: students revealed more genuine motivation in those first nine questions than in any later interaction. Students using the 3D environment kept exploring careers they would never have searched for, a pattern one user called productive wandering: bumping into possibilities instead of filtering them out. That finding became the founding premise of the next tool, Synapse. And the journal proved most useful to students who came back at least three times: one visit produced surface-level entries; repeated visits produced entries that referenced earlier thinking, named contradictions, and tracked priorities as they shifted. DartWorld is still running at Dartmouth's Center for Career Design with hundreds of undergrads having used it, and two of its components (the onboarding quiz and the journaling pinboard) have been extracted as standalone tools now used in other institutional programs.

A workshop and digital tool at Dartmouth's Center for Career Design that teaches students to compress messy experience into a story they can actually tell. The premise is that students don't lack experience; they lack a structured method for finding meaningful patterns inside it. Where conventional career services jumps straight to resume polishing or interview scripts, the workshop sits one layer earlier: it helps students discover what their story is before they try to communicate it.

The workshop runs in five phases: a 90-second brain dump in response to "tell me about yourself," compression of that paragraph down to a sentence and then to a single motivational word, card sorts (Knowdell Values, CliftonStrengths, Motivated Skills) to ground that word in evidence, an AI pass that compares self-perception against resume artifacts and surfaces a possible alternative word, and a rewrite. Compression itself is the engine: cutting from abundance produces clearer, more personal results than building from scratch, and the misalignments between how students see themselves and what their resume documents become the most productive conversations in the room.

The same instrument serves two distinct audiences. For upperclassmen, the chosen word becomes a brand anchor that organizes a body of experience; for first-years and direction-uncertain students, the same word functions as a working hypothesis that guides exploration. Both groups report the same outcome: more control over their own narrative, less pressure to have it all figured out. The argument is methodological: career advising is most useful when it offers a repeatable framework students can return to as their experience and goals change, not a single document optimized for a single moment. We are piloting at the Center for Career Design now, with a live web companion and Spring 2026 expansion planned beyond the center.

This project grew from years spent on both sides of the profession: practicing architecture, advising students, and teaching in the studio. Through those experiences, it became clear that students and young professionals needed more than a syllabus or a single advising session. They needed a place to prepare, to be supported, and to access the practical knowledge that usually gets passed down informally.

Threshold is a free career toolkit for architecture students and emerging professionals. It includes an interactive firm search map with over 2,000 verified firms across all 50 states, guides for writing cover letters, building resumes, and developing portfolios, a step-by-step application timeline, and curated industry resources. You can search firms by discipline, location, and specialty, filter for internship programs, and go directly from discovering a firm to preparing your application.

Every part of the site is built to be useful whether you are applying to your first firm or planning a next move. The guides are specific to architecture. The firm data is verified. The advice comes from real experience advising students through the process.

A career exploration tool that grew directly out of DartWorld's productive wandering finding: students surfaced their most useful discoveries when they bumped into careers they had not been searching for. Synapse builds that pattern into its structure. Instead of returning a ranked list of suggestions, it returns a network of related careers and asks the student to move between them, deciding which connections are worth following.

The interaction is a sequence of small, scaffolded prompts: binary choices about what energizes you, multi-select questions about the subjects that matter to you, results assembled by AI that interprets responses without prescribing them. Authorship stays with the student. The AI's job is to widen the field of view, not to narrow it on the student's behalf.

As Program Manager, I lead the design and deployment of a semester-long cohort-scale career deliberation programme at Dartmouth's Center for Career Design. The program runs as a research cycle each term: it begins with reflection (a studio where students externalize what they already know about themselves onto cards, prompts, and sketches), moves through interpretation (sorts and pattern surfacing that turn the raw material into a working hypothesis), then visualization (spatial maps that hold the hypothesis next to alternatives without collapsing into a ranked list), and ends in action (a commitment exercise that makes the next concrete step legible).

The six-tool arc integrates Identity Mapping Studio, Tile Sorting, Pattern Distillation, OneWord, Launch Studio, and The Sequence. I direct a team of 20+ student interns across product design, branding, software, and UI/UX tracks, working in partnership with Dartmouth's DALI Lab on the digital infrastructure that carries the program between sessions. The Tools section above (DartWorld, Synapse, Narrative by Design) collects the digital prototypes that have spun out of this program; the program itself is the longer-running experiment they are part of.

A semester-long online graduate seminar at Kent State University's College of Architecture and Environmental Design (ARCH 66995, Spring 2026) that guides architecture students through the development of an academic portfolio. The course frames the portfolio as a curated academic argument: a coherent statement of intellectual position assembled for a defined audience under real time constraints, rather than an archival record of work completed. Its three-part architecture (Narrative → Grid → Production) walks students through that argument across fifteen weeks of weekly deliverables, structured peer-and-instructor critique, and a documented revision practice.

I built the companion site at thresholdarch.com/portfolio-as-narrative because a once-weekly fifty-minute session is a fragile delivery mode for a craft that needs repeated reference: students miss class for studio reviews, live discussion outpaces note-taking, and the practical decisions of portfolio design (grid logic, type pairings, when to break a system) are the kind of material people need to look up the night before a deadline. I wanted the course to be accessible across time, learning preference, and circumstance, and to support students who learn in different ways. The site's structure draws on Universal Design for Learning (UDL) and contemporary research on cognitive load, distributed practice, and self-regulated learning. Material is offered in multiple representations (text chapters for slow reading, a structured table of contents for navigation, case studies as worked examples, exercises for generative practice) so students can engage through whichever modality serves them best. Weekly deliverables distribute cognitive effort across the term so feedback can land before the next decision is made, aligned with the well-established research finding that spaced practice produces deeper transfer than massed effort. Revision logs externalize students' thinking back to them, supporting the metacognitive habit of treating revision as the work itself rather than the cleanup after it.

The thread running through the course is curatorial: the most useful skill in portfolio design is choosing what not to include, deciding what each spread is arguing, and aligning every visual decision to that argument. The seminar embodies that thesis at every scale: each session is a checkpoint, each draft is input to critique, and the companion site carries the structural backbone forward so students can return to the workshop chapters, case studies, and exercises long after the seminar's fifteen weeks have ended.

I founded and led Studio Looper LLC, a consulting practice that ran more than forty intensive workshops on academic planning, reflective writing, and portfolio narrative for design-school applicants. The workshops travelled across the United States, Korea, China, the Middle East, and Southeast Asia, with recurring formats including Architectural Boot Camp in Shanghai and the Metamorphic Workshop in Shenzhen.

The pedagogical thread was the same one I carry into the teaching work now: design school applications are most useful when treated as a curatorial exercise rather than a credential-collection exercise. Students arrived with a folder of unsorted work and left with a portfolio that argued for a position. The workshops sat at the seam between the studio crit and the application deadline, treating the portfolio as a working document that had to be defended in front of peers before it was defended in front of an admissions committee.