ABOUT
Hello, I'm Thomas Knoepffler, a design technologist in Miami, Florida. I design physical objects and software tools that respond to the people and spaces around them, working with AI, sensors, and digital fabrication. My recent projects include a 3D-printed hydroponic planter, a desk lamp that senses posture, and a browser tool for building procedural 3D worlds.
I hold an MS in Design Technology from Cornell University and a BS in Integrated Design & Media from NYU, and before graduate school I worked in UI/UX design and product management at an early-stage startup. Outside of studio work I have mentored student designers and volunteered with local creative communities.
I'm currently looking for design technologist and prototyping roles in Miami, Florida, and I'm glad to hear from anyone about the work.
SKILLS
— Rhino 3D, Grasshopper, Blender
— Adobe Creative Cloud, Figma
— 3D Printing, Laser Cutting
— Arduino, Raspberry Pi
— HTML, CSS, JavaScript
— Python, C#
LINKS
Resume
LinkedIn
Instagram
GitHub
Email MIXI II
MIXI II
Cornell University, Cornell Tech
Interactive Device Design, Fall 2025
— INFO 5345
COLLABORATORS
Carrie Wang, Julia Chen
MY ROLE
Designer, Prototyper, Fabricator, Assembler
DESCRIPTION
A countertop mixology machine that listens to how someone feels and pours a tea-based drink to match, choosing from a recipe library and dispensing each ingredient by pump.
LINKS
GitHub
WHAT AND WHY
MIXI II makes a drink from how someone feels. The first MIXI had already mapped emotions to recipes, but it could not do so in real time, so this version was rebuilt from the ground up. A CHI 2025 paper on emotion-based mixology suggested the ingredient: teas, chosen for their effects on mood. Tell the machine you are stressed, or happy, and it chooses and pours a tea-based drink to match.
PROCESS
Feedback on the first version pointed to the social side of mixing a drink, so the redesign began with storyboards and user flows for a shared setting, with existing beverage dispensers as reference. The emotion-to-recipe mapping went into a YAML file, the pumps were calibrated, and the interface was mocked up with a rotary encoder and a small display. Once early tests confirmed that the electronics and the fluid path worked together, the housing was modeled in Rhino and Grasshopper and laser-cut from balsa. More than a dozen people tried it at a showcase, and their comments on the menu and the recipes led to two changes: simpler navigation and ambient music over Bluetooth.
HOW IT WORKS
Two boards share the work. A Raspberry Pi 5 runs the interface on an ST7789 display, listens through a USB microphone, transcribes what it hears, and sends the text to the OpenAI API for sentiment analysis; a Python script then picks a recipe from the YAML matrix and scales it to a fixed 148 mL pour. It passes the result over serial to an Arduino Pro Micro, whose firmware drives six peristaltic pumps through a relay board at a calibrated 21.9 mL per second each. The balsa enclosure is glued, sanded, and routed internally for tubing so that every pour lands in the same place.
NEXT STEPS
The next version would listen more closely, weighing individual words in a conversation to handle mixed or complicated feelings, and would generate new recipes rather than choosing from a fixed set. A companion app could let people share a drink remotely, and a touchscreen or fuller voice control would replace the encoder. A modular build would open the machine to other liquids and to bars and cafés. Two threads of testing continue regardless: whether the drinks actually taste like the moods they are meant for, and how to analyze sentiment without storing what people say.