ABOUT
THOMAS KNOEPFFLER
thomknoe@icloud.com

Thomas Knoepffler is a Design Technologist based in Miami, Florida. His work explores the intersection of materials, interactions, and environments using emergent technologies—such as AI, responsive systems, and digital fabrication. His work spans across mediums, ranging from parametric bio-hybrid products to ambient ergonomic devices to 3D procedural world builders.

He holds a MS in Design Technology from Cornell University and a BS in Integrated Design & Media from NYU, as well as professional experience in UI/UX design and product management for early-stage startups. He has mentored aspiring designers and volunteered with local creative communities alike, dedicated to making design matter for the tomorrows to come.

  • Rhino 3D, Grasshopper, Blender
  • Adobe Creative Cloud, Figma
  • 3D Printing, Laser Cutting
  • Arduino, Raspberry Pi
  • HTML, CSS, JavaScript
  • Python, C#


Resume
LinkedIn
Instagram
GitHub
MIXI II

MIXI II

 Cornell University, Cornell Tech
 Interactive Device Design, Fall 2025
— INFO 5345


DESCRIPTION

An AI-integrated, automated mixology device that uses a RAG-based recipe-driven framework to generate various beverage permutations based on user sentiment and select liquids, measured for ideal refreshment.
MIXI II translates emotional inputs into tailored beverages for homes and social gatherings. The earlier version mapped emotions to drink recipes but lacked feasibilty for real-time implementaion. Inspired by a CHI 2025 paper on emotion-based mixology, MIXI II uses teas for their mood-enhancing properties. It crafts refreshing drinks that respond to emotional states such as happiness or stress and reflect human sentiment.

Development incorporated feedback from the first iteration to strengthen the social dimension of the mixology experience. The process combined storyboarding, user flow mapping, and inspiration from existing beverage dispensers. Emotions mapped to tea-based recipes in a YAML file. Pumps calibrated for accuracy, and the interface prototyped with a rotary encoder and compact display. Early testing verified electronics integration and reliable fluid flow. Laser-cut balsawood enclosures formed the housing using Rhino and Grasshopper. User testing with over a dozen participants at a showcase delivered feedback on interaction clarity and recipe appeal. Refinements added clearer menu navigation and ambient music via Bluetooth. This rapid prototyping cycle produced an intuitive and approachable device.

MIXI II integrates hardware and software for seamless operation. A Raspberry Pi 5 manages the user interface, speech recognition through a USB microphone, and OpenAI API calls for sentiment analysis. An Arduino Pro Micro controls six peristaltic pumps via a relay board for precise liquid dispensing. Recipes store in a YAML matrix and scale to a fixed 148 mL volume based on calibrated flow rates of about 21.9 mL per second per pump. A Python script on the Raspberry Pi handles menu navigation with the rotary encoder and ST7789 display, transcribes audio input, and selects drink mixes. Arduino firmware processes serial commands to execute the pumping. The laser-cut balsawood enclosure assembles with adhesive, sands for a clean finish, and integrates tubing for reliable liquid routing and consistent output.

MIXI II offers strong potential as a social mixology device. Future iterations enhance the AI to analyze individual words in conversations for more nuanced recipes and complex emotional states. The system adds custom recipe generation, companion apps for remote sharing, and improved touchscreen or voice controls. Modular designs support different liquids and commercial environments. Continued testing refines taste mappings and addresses privacy in sentiment analysis. The device evolves into a product line that transforms emotional expression into shared drinking experiences. It blends technology with social ritual to foster deeper human connection.