AfyaQuest

figma file

Inspiration

Community Health Volunteers (CHVs) are the EMT-equivalent of rural Kenya: unpaid volunteers who are the first responders in medical emergencies. But their training today is outdated —delivered through six-hour lecture blocks stretched over two weeks. The result: poor knowledge retention, inconsistent skills, and no way to capture real-time data from the field.

What it does

AfyaQuest transforms boring lectures into engaging, bite-sized learning experiences that CHVs can access remotely. Our main features include: Gamified micro-learning: daily quizzes, short videos, and progress tracking that keep training engaging Map interface: shows visited and pending households for better workflow management Daily reporting & weekly feedback: creates the first systematic way to track health data across communities AI-powered chatbot: allows CHVs to ask medical questions on scene and receive instant guidance

How we built it

Our backend is driven by Node and Express which handle authentication, user system, learning management, AI chat assistant, and more. Our frontend was designed first with Figma and then coded with React and Typescript. We then used MongoDB to handle user management and educational content. We also used the OpenAI GPT-3.5 for our chatbot.

Challenges we ran into

We ran into some challenges with the login page. Sometimes, it would get stuck on loading and never load the next page, and sometimes it would simply consistently return errors. This made it hard as we couldn't check our edits to the other pages without passing the login page. We also ran into trouble when deciding an API to use. We ended up using an API that a team member already had and implemented it into our chatbot.

Accomplishments that we're proud of

We added sounds into both the opening of the website and the quizzes. Although it was a random detail, we were proud of figuring out how to add sounds into the website for an added element. Additionally, we're proud of the map integration and the working keys on the map. We were specifically focusing on a region of Kenya and had to implement the hospital and clinics in the area as well as ensuring that the map would initialize on the correct area.

What we learned

This was the first hackathon or project for most of us. So, we learned lots about the general workflow and structure of how these projects are structured and how to work on it together by designating different parts to different people.

What's next for AfyaQuest

In the future, we'd add offline AI capabilities, more language support, and integration with local health systems. The additional language support would aid in expanding this program to different areas where it could also be applied. When integrating with local health systems, we can get further information about what is 'normal' or 'expected' for different areas to discover any surprising trends that can possibly predict future outbreak.