Skip to content

Work

ALLCOMPUTATION & RESEARCHDESIGN & PRACTICEEXPLORATIONS
THE BOOK (PDF)

21 projects

The Thoughts

  1. behavior information modelingJAN 2026For thirty years we've been getting very good at Building Information Modeling: every beam, every duct, every clash, tracked to the millimeter.
  2. neuroaestheticsDEC 2025Neuroaesthetics is the unglamorous name for a question everyone already feels: why does one room settle you and another keep you on edge?
  3. physics solversAPR 2026A physics solver is the most honest collaborator I have: it doesn't care what I intended, only what I actually built.
  4. generative aiJUN 2024Generative AI made the inspired gesture free. Anyone can produce a thousand renders before lunch, and a render is a promise, not a product.
  5. extended realitySEP 2022I started in a lab strapping headsets onto students to see whether a lesson lands harder when you can walk around inside it.
  6. comfort as dataMAY 2026You don't walk into a room and average your experience. You walk in and the thing that's wrong is the thing you notice: the glare, the echo, the cold draft on the back of your neck.
  7. drawing as interfaceMAR 2026For most of my life the drawing was the output. You designed the thing, then you drew it to prove it existed.
  8. evolutionary searchNOV 2025Evolutionary search is what I reach for when a problem is too tangled to solve head on: describe what good looks like, then let a population of designs breed, mutate, and compete until something clever falls out.
  9. heritage meets new techAPR 2026The usual choice with an old building is a trap: treat it as a museum piece no one may touch, or drop a glass-and-steel box in the middle and call it modern.
  10. buildings that respondJUN 2026Right now a building is a fixed guess. Someone chose the ceiling height and the window size years before you arrived, and you live inside their averaged assumptions whether they fit you or not.
  11. connecting the dotsJUN 2020When the art class was asked who here dislikes drawing, mine was the only hand that went up. Awkward.
  12. pelagñouJUN 2026Some projects hand you a tool; Pelagñou handed me a library card.
  13. explaining thingsJUL 2026How do you explain a complex, honestly boring technical subject so someone leans in instead of glazing over?
  14. adjacency is not accessJUN 2026A floor plan tells you which rooms touch. It will not tell you which rooms you can actually reach, and the gap between those two drawings is where the architecture is hiding.

Sensi

RESEARCHMACAD '26 WINNER

When you fix one sense in a floor plan, what happens to the other five?

Comfort, designed on purpose: a copilot scores a plan across six senses, calibrated to a person, not an average.

MACAD STUDIO · TEAM OF 4

PYTHON · LANGGRAPH · FASTAPI · REACT · LLM BENCH · 2 PROVIDERS × 3 SCENES

BEHAVIOR INFORMATION MODELING ›BLOG (opens in new tab)GITHUB (opens in new tab)
1 / 11

WHAT

Every tool in the stack could tell us how a building performs. None of them would say how a room feels. Sensi closes that gap: a copilot that reads a floor plan and scores comfort across six senses (thermal, visual, acoustic, spatial, olfactory, tactile), calibrated to one person at a time, not an average. Project lead, A to Z, built with a team of four: Lakzhmy Mari Zaro, María Sánchez Domínguez, Charles Abi Chahine, and me.

WHY

Comfort is usually the thing we hope shows up after the design is done. I wanted it to be a layer you can interrogate while the plan is still soft, because you do not walk into a room and average your experience: the thing that is wrong is the thing you notice.

HOW

  1. Onboarding calibrates the copilot to the person in the room, their thermal grudges, their noise tolerance. Comfort without a subject is just weather.
  2. One action classifier, a single LLM call per turn, routes each request through a LangGraph state graph: analyze, edit, preview, audit.
  3. A coupling matrix ripples every change into the neighbouring senses, so a fix that quietly breaks another score gets flagged, not hidden.
  4. The copilot suggests edits the layout can absorb, previews them without committing, then renders the comparison and hands over the report.

WHAT CAME OF IT

We benched the judgment before trusting it: two LLM providers scored the same three scenes end to end, including a living room arranged to fail. They mostly agreed, and it would have been easy to call that validation. We wrote agreement is not truth into the notes instead, and kept every disagreement as data.