Inspiration

When brainstorming project ideas, our team believes in making essential workers' lives easier. This helps them focus on what truly matters. We loved the mission that Impiricus aimed to accomplish. However, we realized a critical field of doctors was left out. Many solutions are built for human healthcare, leaving our animal friends—and their caregivers—behind. A veterinarian’s job doesn’t end when the final patient leaves. It ends after countless hours of staying current on new treatments. Even then, with new drugs hitting the market regularly, it’s hard to keep pace. So why not apply dynamic and critical, on-demand access to vets? That’s where NotiVet comes in.

What it does

NotiVet is a veterinary pharmacy notification application and database that connects clinics and doctors to pharmaceutical manufacturers. We deliver timely and curated drug updates to Veterinarians who can subscribe to their specialty through the website and email. Manufacturers have their own dashboard to create “notification campaigns” that allow them to advertise their new drug and send alerts to the appropriate veterinarians. They are able to see the analytics of how well their outreach campaigns are working, and interested doctors can send emails to the corresponding manufacturer. Scooby, the AI-powered chat assistant trained to specialize in animal medicine, is available to answer any questions one may have and provide accurate, source-backed information.

How we built it

We started by identifying the problem of connecting veterinary clinics and manufacturers and sketching out the workflow of our solution - from user onboarding to personalized notifications. We wireframed the screens required and sketched them out to ensure each step was as intutive as possible. Then, we divided the workload between us based on our personal strengths and chose our tech stack. We implemented Warp to help supercharge the development progress, and conducted mini sprints and reconvined every so often to discuss our progress and next steps. Once we got closer to finishing the project, we centralized our workflow through a GitHub repository.

Challenges we ran into

The first major challenge we encountered was finding a data set for Animal drugs and their purpose. We started by going to trusted sources like the FDA, EMA, and Health Canada. We wanted to have accurate information on approved drugs since our topic is one where a singular error can lead to disastrous consequences. Eventually, after scouring around, we were able to find a CSV containing all the officially approved FDA animal drugs and imported it into our database.

The second challenge was designing the AI chatbot. We wanted the information of the answers to have their sources come from the database, but also wanted to leverage the data training of LLMs to provide additional context to what the user was asking. We decided to use OpenAI’s GPT-4o that was trained on the database of vaccines and supplements we gathered, and linked each reponse to a set of recommended drugs for the occasion. To make the chatbot more personable and approachable, we gave it the name Scooby after everyone’s favorite investigative dog.

Our third and final challenge was designing and creating the backend database: the backbone of our application. It not only needed to contain our data set, but also securely store our users and their passwords, and additional drugs that pharmaceutical manufacturers created in their notification campaigns. We switched our database management system three times, first implementing a NoSQL solution with MongoDB. We then switched to SQLite; however, we encountered limitations with compatibility and ease of use. We finally settled on PostgreSQL, which offered the best balance of features, ease of use, and security.

Accomplishments that we're proud of

We are extremely proud of the simplicity and intuitiveness that our application promotes! The UI is not only sleek and modern, but it also allows for veterinarians and pharmaceutical manufacturers to easily navigate the website and perform whatever they need to do. Our AI assistant Scooby is another feature we have confidence in: its ability to apply what it knows to the specific scenarios from veternarians is a game changer!

What's next for NotiVet

Federations: We want in the future for NotiVet to support hierarchies and company-wide settings, along with single sign-on. Mobile App: We want to create an iOS and Android app that allows veterinarians to easily check new drugs from the comfort of their phone!

Share this project:

Updates