SnakeAid

Instant Snakebite Guidance, Right at Your Fingertips.

Comment

SnakeAid - NLP-Powered Snakebite Assistance for Rural Communities

Hack The Globe 2025 Submission

Inspiration

Snakebite envenomation is a serious yet overlooked global health issue, especially in rural areas where medical assistance is delayed. Inspired by the need for instant, reliable first aid guidance, SnakeAid provides emergency recommendations via WhatsApp, ensuring accessibility even in low-connectivity regions.

What does our solution do?

SnakeAid is an AI-powered bot that helps with:

  1. Fast Stabilization & Transport: Guides immediate first aid actions to get victims to medical care quickly.
  2. Right Decisions Matter: Prevents mistakes, reducing risk of severe complications and therefore likelihood of disability or death.
  3. Lowers technical access barriers: Delivers lifesaving guidance in remote areas without the need of smartphones or internet.

Innovation: Unlike existing first-aid websites or static databases, SnakeAid is instant, interactive, and accessible offline via WhatsApp. No current WhatsApp-based AI snakebite assistant exists for rural users. Google Search is unreliable, and health apps require stable internet & large downloads.

How It Works (Technical Flow)

  1. User sends a message via WhatsApp describing a snakebite incident.
  2. The bot asks guided questions to extract key details (symptoms, location of snake).
  3. Natural Language Processing (NLP) matches the description against the database.
  4. The bot provides first-aid steps to decrease risk of harmful aid.

How We Built It

  1. Backend: Flask (Python) API
  2. Messaging Service: Twilio’s WhatsApp API
  3. Machine Learning: NLP for symptom matching (fuzzywuzzy)
  4. Database: Predefined structured dataset on venomous & non-venomous snakes
  5. Deployment: Hosted via Ngrok (for testing), scalable to AWS/GCP for production

Challenges We Faced

  1. Twilio Restrictions – Limited free-tier messaging required workarounds for testing.
  2. Real-Time Response Optimization – Ensuring quick responses under WhatsApp’s API rate limits.
  3. User Flow Complexity – Iterated multiple times to make guided questions simple and effective.

Overcame these by:

  1. Using session tracking for multi-step conversations
  2. Implementing efficient fuzzy matching to balance accuracy & speed

Accomplishments & Learnings

Successfully built an MVP in under 48 hours! Integrated WhatsApp NLP-based bot with real-time first aid responses. Developed a structured approach for symptom-based venom classification. Improved data efficiency for low-bandwidth users in rural areas.

What’s next?

Short-Term Improvements:

  1. Analyse snake pictures and reply with species information and first aid tips.
  2. Improve NLP accuracy with better entity recognition.
  3. Optimize WhatsApp API interactions for faster response times.

Long-Term Scaling:

  1. Deploy on cloud infrastructure (AWS Lambda, Firebase) for wider adoption.
  2. Expand treatment to other regions and neglected tropical diseases such as spider bites and plant poisoning.
  3. Partner with hospitals, emergency services & NGOs to integrate real-time response alerts.
  4. Monetization: Offer B2B partnerships with hospitals & insurers, with API-based pricing.
Share this project:

Updates