FieldReady

Field Ready Patient Simulation
Field Ready Flashcards
Field Ready Response Area Quiz
Field Ready Radio Simulation

Inspiration

EMT training also requires learning response areas (specifically street names and numbers), so trainees know where to go when dispatched. We added a map simulation to help trainees learn this.

What it does

FieldReady provides four training modules. The Patient Simulation generates random EMT scenarios with patient vitals. Trainees interact with AI-powered patients using voice or text, ask assessment questions, and receive realistic responses. The AI dispatcher provides feedback on assessment technique and guides trainees through proper protocols like ABCDE and SAMPLE. The system generates vital signs based on the scenario and includes voice synthesis for dispatcher and patient dialogue.

The Response Area Quiz uses interactive maps to help trainees learn building locations and addresses in their response area. Trainees identify buildings on a map, reinforcing geographic knowledge needed for dispatch.

The Flashcards module lets trainees study EMT protocols, medications, dosages, contraindications, and vital sign ranges. Users can create custom decks and review key information.

The Radio Simulation module will let trainees practice radio communication protocols and dispatch procedures.

How we built it

We built FieldReady with a React frontend and an Express.js backend. The patient simulation uses Google's Gemini API for script generation and line-by-line dialogue. We integrated ElevenLabs for voice generation, creating realistic dispatcher and patient voices that adapt to age and gender. For the map feature, we used Leaflet.js to create an interactive mapping system that pulls real building data and helps trainees learn geographic locations. The frontend uses TypeScript, Tailwind CSS, and modern React patterns for a responsive, accessible interface.

Challenges we ran into

Integrating ElevenLabs for voice generation was challenging. We had to handle API rate limits, manage audio playback states, and adjust voice parameters based on patient characteristics and emotional state (using stage directions from the AI-generated scripts). We also worked through merge conflicts while collaborating, which improved our Git workflow and team coordination.

Accomplishments that we're proud of

Using a voice AI API for the first time felt like a step forward, hearing the AI-generated dispatcher and patient voices made the simulation feel more real. We're proud of the architecture that combines Gemini for script generation with ElevenLabs for voice synthesis, creating a cohesive training experience.

For me (Will), the map generation was especially rewarding. During the summer, I never got the chance to make a local area settera clone (map quiz site) like I wanted to at the beginning of the summer. Building the Response Area Quiz for FieldReady felt like an extension of that idea, and seeing it come together as a practical training tool was fulfilling.

We also improved our collaboration skills by resolving merge conflicts and learning to work effectively as a team.

What we learned

We learned to manage merge conflicts and improve our Git workflow. We gained experience integrating multiple AI APIs (Gemini and ElevenLabs) and handling asynchronous operations, audio playback, and state management in React. Working with Leaflet.js expanded our mapping skills, and building a training application deepened our understanding of user experience design for educational tools.

What's next for FieldReady

We plan to expand the scenario library with more diverse medical emergencies and add difficulty levels. The Radio Simulation module will be fully implemented with realistic dispatch scenarios. We want to add progress tracking and performance analytics so trainees can monitor improvement. Multiplayer scenarios where multiple trainees work together on the same case would add realism. We're also considering mobile app versions for on-the-go training and integration with official EMT certification programs to align with standardized curricula.

https://github.com/mrbumcum/dispatch