UnifyMD

Title Slide
Problem
Innovation
Key Features

🌟 Inspiration

Long lists of patient records make it challenging to locate relevant health data. This can lead to doctors providing inaccurate diagnoses due to insufficient or disorganized information. Unstructured data, such as progress notes and dictated information, are not stored properly, and smaller healthcare facilities often lack the resources or infrastructure to address these issues.

💡 What it does

UnifyMD is a unified health record system that aggregates patient data and historical health records. It features an AI-powered search bot that leverages a patient's historical data to help healthcare providers make more informed medical decisions with ease.

🛠️ How we built it

We started with creating an intuitive user interface using Figma to map out the user journey and interactions.
For secure user authentication, we integrated PropelAuth, which allows us to easily manage user identities.
We utilized LangChain as the large language model (LLM) framework to enable advanced natural language processing for our AI-powered search bot.
The search bot is powered by OpenAI's API to provide data-driven responses based on the patient's medical history.
The application is built using Next.js, which provides server-side rendering and a full-stack JavaScript framework.
We used Drizzle ORM (Object Relational Mapper) for seamless interaction between the application and our database.
The core patient data and records are stored securely in Supabase.
For front-end styling, we used shadcn/ui components and TailwindCSS.

🚧 Challenges we ran into

One of the main challenges we faced was working with LangChain, as it was our first time using this framework. We ran into several errors during testing, and the results weren't what we expected. It took a lot of time and effort to figure out the problems and learn how to fix them as we got more familiar with the framework.

🏆 Accomplishments that we're proud of

Successfully integrated LangChain as a new large language model (LLM) framework to enhance the AI capabilities of our system.
Implemented all our initial features on schedule.
Effectively addressed key challenges in Electronic Health Records (EHR) with a robust, innovative solution to provide improvements in healthcare data management.

📚 What we learned

We gained a deeper understanding of various patient safety issues related to the limitations and inefficiencies of current Electronic Health Record (EHR) systems.
We discovered that LangChain is a powerful tool for Retrieval-Augmented Generation (RAG), and it can effectively run SQL queries on our database to optimize data retrieval and interaction.

🚀 What's next for UnifyMD

Partnership with local clinics to kick-start our journey into improving healthcare services and patient safety.
Update to include speech-to-text feature to increase more time patient and healthcare provider’s satisfaction.