SimuCare

Inspiration

SimuCare was inspired by a real, urgent problem in healthcare education: the gap between theoretical knowledge and practical readiness in healthcare settings, especially in under-resourced environments. In addition to that, traditional simulation tools tend to be rigid, fixed, and often expensive, creating a barrier to many students seeking to augment skills on their own pace. It was born out of a frustration for how limited the available options are to students who may not have access to costly, high-fidelity simulations. We wanted something that could mimic the pressures and nuance of clinical situations but were customizable to what we are learning at the moment.

What it does

In essence, SimuCare allows users to practice clinical scenarios in a controlled, virtual setting. Scenarios are either generated based on the course material the user provides (in image or text form), or are scenarios that are innate to the app itself. The app allows for instant scenario creation, evidence-based clinical situations, access to hundred of pre-built scenarios across all nursing specialities, AI agents that simulate patient interactions, real-time feedback, immediate error correction, learning path suggestions, difficultly adjustments, dynamic patient conditions and personalities and lastly, performance analytics. Progress can be tracked once logged in, enabling weakness identification and personalized study plans.

How we built it

The main framework we chose to experiment with while building the website application was TanStack Start. TanStack is a powerful framework that makes building modern websites much more efficient and robust. It possesses powerful tools including TanStack Query and TanStack Router that help with tasks like efficient data fetching and managing routing, so we can focus more on building features instead of worrying about repetitive setup. Externals API's were also integrated such as Gemini for building A.I powered features. Supabase is use as our main database to store and manage data as well as for authenticating users.

Challenges we ran into

Originally, we had a vision for a more conversational experience that was the closest thing you can get to the timing of real-world conversations and dialogue. We were even thinking of integrating real-time AI video agents to add to that immersive feel and make the simulation meaningful and stick to the user. However, due to time constraints, we opted out for a more chat-based prototype that we believe can grow from there.

Accomplishments that we're proud of

We are personally proud of getting started with, and working towards, subverting the narrative that AI will eventually replace much-needed healthcare roles, or that a reliance on AI only does harm. We are proud of coming up of an application that showcases an apt use of AI for the supplementation, and not the substitution, of the healthcare field. Additionally, we got to the level that we are currently in our respective journeys with software development skills given the time constraints via concerted efforts to flesh our ideas out into tangible code.

What we learned

As a team, we learned to navigate the complexities of all the platforms we used, utilize our time and embrace the diverse set of skills that each individual member of the team has to offer for coordination purposes. For the software engineering students of the team, valuable skills on how to work with and implement AI beyond what is taught in the classroom. For the nursing student in the team, relevant skills in software engineering were explored, and the interplay of nursing and software engineering became well-acquainted.

What's next for SimuCare

Refining the simulation engine and coaching system is our next step in order to provide more accurate and diverse training. Continuation of the UI improvement, realism and possibly uploading features are areas of improvement that we think are necessary avenues to embark on.