Sophomore Vignav Ramesh blearily looked down at the rows of multicolored code. It was 2 a.m., and his group had only seven hours left to perfect their final projects. With the pressing deadline, they began frantically debugging their code and bouncing ideas between themselves.
Ramesh, along with sophomores Rohan Chakravarthi and Ayaan Haque, participated in the Omnihacks hackathon in San Francisco on Sept. 22 among 35 teams and 130 competitors. The three designed Tickbird, a text-to-speech app allowing visually impaired people to understand prescriptions from their doctors.
Tickbird, named after birds that guide rhinos through the wild, was inspired by Ramesh’s uncle, who is rapidly losing his vision.
“Watching him struggle with everyday actions that we all take for granted, I was concerned about his vision’s impact during more important situations such as understanding prescriptions from his doctor or pharmacy,” Ramesh said.
Through research, the group found the problem stretched far wider than they initially believed. Only 14 percent of pharmacies and 1 percent of doctors adjust their prescription treatments to accommodate visual impairments, drastically increasing the chances of significant mistakes such as overdosing.
The group went to work making plans for the app, first placing themselves in the shoes of Ramesh’s uncle. They envisioned how their app could best aid his lifestyle and needs.
The group built a user profile page, a prescription scanning page and a page displaying a user’s past prescriptions. Users enter their personal information into a shared database, and the prescription scanning page allows them to translate a picture of any prescription into text. Users can then hear their past prescriptions read aloud, bypassing any visual impairments risking prescription misuse.
With their app design in place, the group traveled to San Francisco to attend Omnihacks, where they had 24 hours to build their design from scratch. The team worked tirelessly, taking breaks only to eat and sleep on the floor in sleeping bags.
Haque found teamwork an especially important factor while designing the app, especially during crunch time. The team divided up the tasks and worked side by side, collaborating to connect their respective segments of the app together.
“Teamwork made us much more efficient,” Haque said. “We tailored each part of our app to the strengths that the three of us each have.”
Tickbird placed top 10 overall in the hackathon, also ranking as the second-best mobile app among over 35 teams. The hackathon experience provided the team with a closer look at the burgeoning fields of AI and machine learning.
“A lot of other successful teams used AI/ML in applications regarding computer vision and algorithmic problem-solving, which was really interesting and insightful,” Ramesh said.
The team plans to keep building upon their app, starting with adding a smart profiling feature allowing users to find locations with their specific prescription. They also plan to compete in future hackathons.
“Who knows?” Haque said. “We’re still editing and we think that this same app can be used for a lot of things, maybe even posted on the App Store or given to doctors.”
Demo of app: https://www.youtube.com/watch?v=7230378jFNk
Github Repository: https://github.com/rvignav/tickbird