The prototype offers two conversations: an AI fitness trainer and an AI meal planner.
What I built
I led the hackathon team, built the React interface, and worked on integrating an OpenAI-powered chatbot with prompts for fitness and meal planning.
Our prototype brought two task-specific conversations into a web experience with progressively displayed responses.
Purpose & audience
We wanted to make the first step toward planning a workout or a meal less overwhelming. The prototype used conversational input to suggest a personalized starting point.
Under the hood.
Follow the architecture, one connection at a time.
AIHealth / architecture
The conversation asks for personal context and, for meal planning, cooking experience to inform its suggestions.
Task-specific prompts guide the ChatGPT integration toward workout plans or meal ideas.
The web interface displays the generated response progressively, as documented in the original project write-up.
Scroll to follow the connections. Select a step to take a closer look.
Decisions & tradeoffs
We adapted an existing ChatGPT wrapper and focused on two clear use cases within the hackathon timeframe. Understanding the wrapper and connecting its behavior to the interface was a central integration challenge.
Outcomes & limitations
The team completed a working hackathon prototype. The public write-up and repository document the original implementation and what we learned about integrating React, JavaScript, and AI responses.
This was an exploratory project. A functioning chatbot does not establish the accuracy or suitability of its health-related suggestions, and the original write-up describes public hosting and a mobile app as future plans.
What’s next
Possible directions for further development.
The original team proposed refining usability and design, hosting a public version, and exploring a mobile app. These remain possible directions rather than completed features.