Second-year computer science (CS) majors and entrepreneurs Syed Husain and Vihaan Nagarkar developed MemoLens as participants in DubHacks ’24, an annual student-organized hackathon at the University of Washington.

The duo won first place at the hackathon, held Oct. 12-13, in the AWS Gen-AI and Cloud Computing track.

MemoLens combines smart glasses with facial recognition algorithms and a customizable AI assistant to help Alzheimer’s patients:

• Recognize loved ones
• Keep up with daily medications
• Lead more independent lives

Husain and Nagarkar each had grandfathers who had Alzheimer’s disease. Nagarkar said he was motivated to help when he saw his grandfather struggling to put names together with familiar faces.

“It was heartbreaking to see the confusion and anxiety on his face when he couldn't recognize his own family,” said Nagarkar, who attended high school at the British International School in Lagos, Nigeria.

These recollections inspired Husain and Nagarkar. With their DubHacks '24 project, Empowering Memories, the duo created a platform to help those with Alzheimer’s to retain a sense of independence and stay connected with family and friends.

The MemoLens platform integrates the customizable AI assistant with Vuzix M400 smart glasses. The AI assistant is trained to recognize faces from a friends-and-family photo database and provides users with medication alerts. MemoLens also answers disease-related questions and connects users with patient resources.

Once they had a working prototype, Husain and Nagarkar put their device through rigorous user testing. They used the feedback from healthcare providers and patients to address any challenges and improve the platform.

“Through DubHacks, we learned the importance of empathy in technology design and how critical end-users are in the design process to create solutions that genuinely meet their needs,” said Nagarkar.

Although the platform started as a hackathon project, the entrepreneurs plan to develop MemoLens and bring it to the marketplace.

“We aim to enhance the system by developing a multi-LLM architecture, where a central LLM acts as an agent that seamlessly delegates tasks to specialized models,” said Husain, who graduated from Plano West Senior High in Plano, Texas.

The team expects the integrated multi-model system and upgraded hardware will reduce latency on the platform. This approach will allow MemoLens to offer users more personalization, assistance, real-time health monitoring, and daily reminder support.

“By leveraging specialized models, we can provide deeper contextual understanding and a more tailored experience for each user, elevating our support,” said Husain.

The duo has started researching new wearable tech for MemoLens 2.0. They expect the next iteration to be faster, more responsive, and more impactful. “We want to make solutions an integral part of the tech landscape, using human-centered design to solve complex problems,” said Nagarkar.

“Real innovation happens when ideas are brought to life at scale, creating lasting positive impacts. MemoLens 2.0 is a step toward this vision.”

Husain and Nagarkar won a Smart Home hub with the Echo Show 11 as winners of the DubHacks '24 AWS Gen-AI and Cloud Computing track.