FRANKIE IOVINELLI

SmartSpa was a two-year effort to engineer an autonomous hot tub chemistry management system from the ground up. The final system went well beyond a typical IoT prototype, combining a commercial-grade custom PCB, regulated power architecture, embedded firmware, imported chemical sensors, flow meters, multiple coordinated dosing pumps, fluid handling, wireless connectivity, cloud database hosting, and a fully developed mobile app into one integrated product. The work ultimately led to a patent application covering the system and its approach to autonomous chemical maintenance.

Much of the engineering challenge was in making those mechanical, electrical, chemical, embedded, cloud, and software systems behave reliably together, then iterating the hardware, firmware, calibration, control logic, and data flow until the complete system functioned as a cohesive product.

DesktopAerospace became a hands-on exercise in designing products for repeatable additive manufacturing. I developed a growing line of aerospace models by refining geometry for printability, tuning orientation and support strategy, selecting materials and process parameters, and balancing surface quality, strength, print time, and cost.

Beyond the parts themselves, I built the surrounding production system as well, including assembly, finishing, packaging, photography, fulfillment, and iterative design changes driven by real customer feedback. The project pushed me to think beyond CAD and treat every model as a manufactured product that had to be both technically well-designed and commercially repeatable.

Recall is built around a simple problem in language learning: you see a new word, understand it in the moment, and forget it almost immediately. The app lets users capture those words as they encounter them and gradually build a personal lexical foundation around the vocabulary they actually use and struggle with.

From there, AI expands that foundation by generating contextual sentences, reinforcing related vocabulary, identifying weak spots, and continuously adapting what the user sees next. Rather than treating vocabulary as a static list of flashcards, Recall turns each saved word into part of a growing, personalized language system that becomes more useful as the learner uses it.