An AI Assistant on Computing’s Edge

Qualcomm Computer Science/Mathematics, 2023–24

Liaison(s): Victor Mai, Satyam Gaba
Advisor(s): Arthi Padmanabhan
Students(s): Tanvika Dasari (TL-F), Josh Jansen-Montoya, Leila Li, Nic Tekieli, Kelly Yang (TL-S)

Qualcomm is at the forefront of enabling low-power inference at the edge through pioneering model-efficiency research. To showcase Qualcomm’s on-edge computing framework, our team was tasked with creating an Android app equipped with an AI question-answering assistant. We leverage Qualcomm’s on-device ML capabilities to create a software framework that supports loading and running multiple ML models in one app. Our project serves as a proof of concept that Qualcomm’s ecosystem of solutions can enable on-device ML inference to be used in a plug-and-play fashion.