TARA stands for Treatment and Recovery Assistant.

We want to help patients recover in the post-operative and treatment phase by giving them a means of constant communication with their doctors and providers. We are building a chatbot that personalizes patient-provider communication and serves as a liaison in helping patients ask what is important to them and get answers on-demand.

What it does

Mobile Health Chatbot and Liaison Engages patients to follow up with treatment plans and adherence. Answers patient questions as chatbot in a chat interface, redirecting to care team when most appropriate. Serves as communication tool for progress updates, treatment updates, appointments, regimen updates for medication/therapy, and follow-ups from provider to patients

Reduces risk factors of re-admission, Reduces costs during transition of care Addresses health outcomes after discharge

How we built it

TARA was created as an iOS app in Swift. We created the chatbot using AIML markup by interfacing with the Pandorabots API. The chat interface interface and data views are self-created and backed up using persistent Core Data. All data in the app is synced to Google Firebase and fetched in real-time to ensure that patients get the most important information when they need it.

Challenges we ran into

Learning how to interface with the Pandorabots API was a challenge since neither of us had much experience using it before.

Accomplishments that we're proud of

We're proud we built a fully functioning hack that serves a a first step in building greater transparency in patient-provider communication.

What we learned

How to create a chat interface from scratch and build a chatbot.

What's next for TARA

Implement SiriKit. Communicate with TARA services using just your voice! Incorporate Natural Language processing and sentiment analysis for detecting emotions. Enable to engage with patients empathetically (ex. Depression, fatigue). Improve image recognition tasks for pictures, videos. Used as assistive tool for provider for diagnosis and recognition.

Share this project: