MediBuddie

MediBuddie

The Highlight

An image-text recognition app based on Google APIs which tells you all the information you care about for a medicine. We use FDA's API to get the information to you.

The Inspiration

Due to the current pandemic, not everyone has access to a doctor. Here's where MediBuddie comes in action. And in a lot of cases, taking the wrong dosage or medicines could be a fatal thing to do, especially for the elderly. The developers live away from their homes and have, many a time, found themselves in situations when they are either skeptic of medicine or don't know what medicine to take for the most common health issues. Talking to the community around them back home, they see that because of the inaccessibility of doctors, many people take medicines of common knowledge on their own. While our app does not suggest medicines based on any symptoms, it does validate the use of a certain medicine.

Additionally, one can also use this app to clear any suspicion they might have regarding a medicine prescription. Later, we'd also be implementing "Nearby doctor consultations" and an option to buy the particular medicine at the "Nearest store".

The Process

The app begins with the disclaimer and the instructions page. On continuing through these, we get to the main part of the application which is the camera page.

Using Optical Character Recognition, the foundation of the app was built. Specifically, the app could read the text out of an image, allowing us to read the contents off the medicine. This then went to text processing API and removed the "weed" out of the text and sent it to the FDA API. This fetched out the information we needed and put it in the presentable form.

Locally, we then store this data into your "search history" for faster access next time. This uses some local memory which is minimized by using references for these searches.

We used flutter and integrated it with Firebase. We've developed this application by using a majority of Google's APIs and other APIs to support the work.

The Challenges

One of the major roadblocks for us was communicating with one another during the event. We tried various methods of audio and video calling, used services like Discord and Slack, but ultimately, we used VS Live Share and VS Live Share Chat to talk and discuss the implementation. It also helped that we were able to look at each other's code in real-time to find any bugs that we overlooked.

Another major hurdle was finding the database to do find the drug. We used several online sources. One of the team members made a Web Crawler (based on Python) create a database so that we could use the app offline. However, after almost implementing it, we decided to scrap that idea and use FDA's API.

The Learning

The team members, both freshman, intended to participate in the hackathon for the learning of it. It is almost trivial to mention that participating and attending webinars were fun, but this project helped us see the "work from home" concept first hand. We got through the challenges of working online, learned a lot about APIs and app development.

Finally, for me, this documentation is a very unique experience. I would just like to take a few characters to mention how this writing is both enthralling and new.

The Business Model

We have created a working application that connects and lets you know about the medicine you have. Now, we can further this implementation to make it a "store" of sorts. When you have a medicine or you search for a particular drug, you'd probably want to buy it. There comes the app. The app connects you to your local medical store and helps you call/place an order with them. We are also looking to add a directory for doctors which one could use to get one time, instant advise or set up an appointment, both online and in-person.

This could create revenue for us by store commissions. We could also charge a small fee to connect to a doctor. Bootstrapping can be done in a small town like Davis. We could also talk to pharmacies and get them on board with this by providing them with an existing user base.

The Road Ahead

In the limited time that we had during the event, the team created a working application and created a business model. However, in the upcoming days, we are looking forward to implementing the business model we have created and looking to improve/redo the UI. It would require us to implement GPS locating, meeting up with the stores, and doctors interested to get on board. We could bootstrap and possibly start talking once the situation around the world normalizes!