Data Science Projects
Health System Project
From March 26 to 30, 2018, I participated in the 3-Day Health Pro Challenge during Penn State’s Startup week.
Responsibilities: Web UI Design, iOS Mock Up, Health Solution Research, Interaction Design, Project Lead
Research & Design Process
General Opportunity
Goal:
The project aimed to find cost-effective solutions for a developing country’s patients during a disease outbreak. It focused on epidemiology, the study of the distribution and determinants of health-related states or events (including disease).
Opportunity:
During epidemics and health emergencies, national infrastructure and health systems can be subjected to stresses that interfere with their ability to provide critical support for citizens. What could help patients to self-diagnose medical conditions and disease?
First, I identified that there is adequate internet and phone access in developing countries. For example, according to the Pew Research Center, in 2015, 81% of Indians and 97% of Chinese had cell phones. The majority of developing countries have 3G Internet coverage.
Target Audience:
Individuals who ordinarily receive little or inconsistent health support from their community or country.
App Design
I designed an iOS mockup that helps patients to self-screen for infection and find venues for medical treatment.

The mockup has two main features: the self-screen and a locator for nearby clinics.
Self-screen: Patients take a ‘smart’ survey, where each question they answer determines
the next, establishing the likelihood that they are infected.
Location: The app searches for nearby clinics.
Additional features
Statistics: The app collects data from local hospitals and shows the number of infected
individuals in the community.
This can be a life-saver; for instance, if the data shows a high level of infection, parents could avoid
cross-infection by limiting their children’s after-school activities.
Advertising: On the app, pharmaceutical companies and drug stores could advertise for
over-the-counter medications.
Feedback from Judges:
Be more explicit on your product, tell the audience what you are providing. It's an App that will collect data on the audience’s symptoms and provide it to local health authorities and NGO's so authorities can respond appropriately.
Give more details on specifically what's in it for the patient, why should they do this?
Responding to the judges’ feedback from Health Pro Challenge(above) I narrowed the general opportunity regarding epidemics and health emergencies down to a more specific opportunity: pet health emergencies. The end result competed in Technica 2018.
Pat Your Pet
Research & Design Process
A Specific Opportunity
Goal:
The project aimed to find cost-effective solutions for pet owners when their pets are sick and needs timely assistance. It borrowed concepts and modified design features from the first project.
Opportunity:
People often have tools to book an appointment with their doctors, such as zocdoc, whereas pet owners don’t have such a tool for their pets.
Target Audience:
Pet owners, especially in medical emergencies
Design Rational:
I designed a web application in R that helps pet owners to diagnose health issues in their pets and find venues for medical treatment, including general insurance information, a diagnostic tool, and a locator for nearby veterinary facilities.

The opening page provides links to pet insurance (not all pet owners have it). Once the owner is aware of this information, he or she can continue to the next tab.

Under the “Profile” tab, the owner enters information about his or her pet, including the type of animal, its physical symptoms, and when the owner would like to meet with a veterinarian.

Last, this application provides a tool to find nearby veterinary facilities. The dollar-sign ranking system is inspired by Yelp.com.
Feedback from Judges:
Most of the judges from Technica 2018 liked the idea of an application for pet owners and found it practical. Most were impressed that the app was built entirely in R, a statistics programming language. The team also received positive feedback for the user interface design. One important observation was that we didn’t have a database connected to the app, so we couldn’t store data or interact with the map in real-time.