Donna Mae Quinto

Four countries and quite a few homes

I grew up in 4 countries, although I've lost count of how many times I've moved. I've called places across Japan, the Philippines, the UAE and Ireland home at different points in my life. Here's the slightly winding route that got me here.

• Nagoya, Japan
• Osaka, Japan
• Manila, Philippines
• Dubai, UAE
• Abu Dhabi, UAE
• Dubai, UAE
• Manila, Philippines
• Shannon, Ireland

Apparently staying in one place was never really my thing.

Pitching to John Collison

When I was 16, I was accepted into a highly competitive summer youth accelerator at Dogpatch Labs in Dublin (Ireland's startup hub). What started as a summer project with 2 college students quickly turned into a startup.

After the accelerator, I presented our work at Demo Day and was later invited to pitch our startup to John Collison at Dogpatch Labs! He gave me incredibly valuable advice as a teenager building a company, and he remains one of the tech figures I admire most.

Still not entirely sure how that happened.

Federated UAV Thermal Anomaly Detection

I built a system where a fleet of UAVs collaboratively learns to detect thermal anomalies such as wildfires and infrastructure faults without ever sharing their raw data. The challenge was that every drone sees a completely different environment, making conventional federated learning unreliable across such non-identical data. I used federated distillation instead, exchanging predictions rather than model weights to make collaboration far more bandwidth-efficient, alongside adaptive loss and uncertainty weighting to handle noisy and corrupted thermal imagery. I validated the approach through non-IID simulations and on a real DJI Mavic equipped with a thermal payload.

Federated Learning · Federated Distillation · Computer Vision · UAVs · Thermal Imaging · PyTorch

Coffee Grinder, with a Brain

I reverse-engineered my coffee grinder and gave it its own grind-by-weight system using an ESP32 and a load cell. The system continuously measures how much coffee has been dispensed and cuts the motor at exactly the right moment instead of relying on a timer. I built the hardware and software around the existing grinder, turning an ordinary appliance into something that could accurately dose coffee on its own. It consistently gets within ±0.03g of the target weight, which is probably more precision than my morning coffee deserves.

ESP32 · C++ · Load Cells · Embedded Systems · Electronics

DiverseForce

I co-founded DiverseForce at 16 alongside two college students after becoming interested in the barriers people with disabilities can face before they even reach an interview. We explored how technology could make hiring more accessible and help organisations think differently about the way they find and evaluate talent. It was also my first experience of building something collaboratively with people from a completely different stage of life and education. The project pushed me to think about accessibility not just as a technical problem, but as something that should shape the entire product.

Web Development · Accessibility · Product Design · UX

Mini Robot Arm

I built a Linux loadable kernel module and character device driver for a Raspberry Pi robot arm. I used it to create a teach-and-replay system where the arm could record movements and reproduce them later using servo control and GPIO interrupts. The project also involved I2C communication, blocking I/O and a terminal-based ncurses dashboard for controlling and monitoring the system. It was one of my deeper dives into how software interacts with hardware below the level of a typical application.

C · Linux Kernel Modules · Raspberry Pi · GPIO · I2C · Servo Control · ncurses