Instant_Firstaid

• 

  running locally on localhost:5173

• 

  Login page

• 

  Signup page

• 

  authentication available

• 

  Homepage

• 

  Explanation of one of the injuries first aid

• 

  A user can download the first aid

• 

  Fast guidance using AI

• 

  profile page

Inspiration

In many rural or underserved communities, first aid knowledge is limited, and emergency response is often delayed. We were inspired by real stories of students, farmers, and family members who didn’t know what to do in emergencies — from fainting in class to animal bites in the field. We wanted to build a simple, offline-friendly AI assistant that could instantly guide anyone through life-saving first aid procedures.

What it does

Instant_FirstAid is a web-based AI assistant that:

Takes in a short description of an emergency (e.g., "someone fainted").

Uses a lightweight AI model (powered by Hugging Face) to generate step-by-step first aid instructions.

Works instantly in the browser, even on mobile.

Offers clear, beginner-friendly advice — no medical training required.

How we built it

Frontend: react and JavaScript (no frameworks, to keep it simple).

AI Engine: Hugging Face Inference API using the flan-t5-large language model.

Challenges we ran into

Choosing a model that gives accurate, short, and medically sensible responses.

Making the app simple enough for use in a real emergency.

Avoiding hallucinations from the language model and handling bad or irrelevant outputs. API calls are done directly in the browser using fetch(), with a Hugging Face API key.

Designed to work with minimal internet and be mobile-friendly.

Accomplishments that we're proud of

Built a working AI-powered assistant in a single day.

Created a tool that could save lives in places where hospitals or clinics are far away.

Kept everything lightweight and fast-loading — no complex installs needed.

What we learned

How to integrate Hugging Face models with plain frontend code.

How to craft prompts to get useful emergency instructions.

How to simplify UI for use under stress — clean layout, fast typing, and minimal distractions.

What's next for Instant_Firstaid

Add voice input so people can speak their emergency.

Train or fine-tune a custom medical model for more reliable results.

Add an offline fallback using JSON for no-internet scenarios.

Expand to include first aid for animals and farmers (snake bites, machinery injuries, etc.)

Convert it into a mobile app using Cordova or PWA.

Built With

• react
• trae