Swasthyashe – Women’s Health & Wellness Platform

Empowering women with holistic health solutions through a fusion of traditional wisdom and modern technology.

Comment

Swasthyashe – Women's Health & Wellness Platform

Inspiration

Growing up in India, I witnessed how my grandmother (Bhamma) was the first doctor in our household — using turmeric for wounds, ginger for colds, and herbal concoctions for period pain. Yet, in the age of modern medicine, this ancestral wisdom was fading away. Only 18% of women in India have access to proper menstrual hygiene, and millions still rely on ash, sand, or cloth during their periods.

Swasthyashe was born from the belief that women's health deserves a platform that honors both ancient Ayurvedic traditions and modern scientific research. The name "Swasthyashe" combines the Sanskrit word Swasthya (health/wellness) with She — health for her.

I wanted to create a space where a woman could track her cycle, understand her body through Ayurveda, get personalized wellness advice from AI, and access curated health content — all in one place.

What it does

Swasthyashe is a comprehensive women's health platform with the following features:

Core Features

  • 🩸 Period Health Hub — Natural remedies for cramps, blood color guide, menstrual hygiene tips, personal first period stories, menstrual clot information, and curated organic pad brand recommendations (Saathi, Azah, Nuva, Vivanion, Bella)
  • 📅 Cycle Tracker — Log periods with start/end dates and flow intensity. Automatically calculates next period prediction, fertile window, and ovulation date using the formula:

$$\text{Next Period} = \text{Last Start Date} + \text{Average Cycle Length}$$

$$\text{Ovulation Day} = \text{Next Period} - 14$$

$$\text{Fertile Window} = [\text{Ovulation} - 5, \text{Ovulation} + 1]$$

  • 📝 Symptom Tracker — Log daily symptoms (cramps, headaches, mood changes, fatigue, bloating) and receive personalized Ayurvedic recommendations with specific herbs, yoga poses, and dietary suggestions
  • 🧘 Wellness Planner — Dosha Assessment quiz to determine your Ayurvedic constitution (Vata, Pitta, Kapha), followed by personalized weekly meal plans, yoga routines, and wellness activities
  • 🤖 Dr. Ayur — AI Wellness Coach — An AI-powered Ayurvedic doctor chatbot built on Amazon Bedrock using Meta Llama 3.2. Ask any question about skin care, period pain, diet, yoga, PCOS, pregnancy, hair care, or stress and get instant, structured Ayurvedic advice with specific herb names, remedies, and practices
  • 👤 User Profile — Editable health profile with personal details (name, contact, email, DOB, blood group, height, weight) with save/edit toggle and localStorage persistence
  • 🔐 Authentication — AWS Cognito-based sign-in/sign-out with user session management
  • 🔍 Search — Search across all health content and topics
  • 💌 Newsletter Subscription — Email subscription for health tips and updates

Health Content Sections

  • 🧠 Mental Wellness — Traditional practices for emotional balance and mental harmony
  • 💗 Sexual & Reproductive Health — Essential information for reproductive well-being
  • ✨ Skin & Hair Care — Natural Ayurvedic beauty secrets and remedies
  • 🤰 Pregnancy Care — Holistic support throughout the pregnancy journey
  • 🍽️ Diet Plans — Ayurvedic nutrition guidance

How we built it

Frontend

  • React 18 with TypeScript for type-safe component development
  • Tailwind CSS for responsive, utility-first styling
  • React Router for SPA navigation
  • CSS keyframe animations for smooth page transitions, card hover effects, and staggered reveals
  • localStorage for client-side data persistence (cycle data, symptoms, profile, dosha results)

Backend

  • AWS Lambda (Node.js 18) — Serverless functions for all API endpoints
  • Amazon DynamoDB — NoSQL database for cycles, symptoms, forum threads, user profiles, and subscriptions
  • Amazon API Gateway — RESTful API with CORS support
  • Amazon Cognito — User authentication with OAuth 2.0 implicit grant flow
  • Amazon Bedrock — Generative AI using Meta Llama 3.2 3B Instruct model for the Dr. Ayur wellness chatbot
  • Amazon SES — Email notifications for newsletter subscriptions
  • AWS CDK (TypeScript) — Infrastructure as Code for the entire stack

Deployment

  • Amazon S3 — Static website hosting for the React production build
  • Amazon CloudFront — Global CDN distribution for fast, HTTPS delivery
  • AWS CDK — Automated deployment of all backend resources

Architecture

User → CloudFront → S3 (React App)
                  ↓
            API Gateway → Lambda Functions → DynamoDB
                                          → Bedrock (AI Chat)
                                          → SES (Email)
            Cognito (Auth)

Challenges we ran into

  1. Bedrock Model Access — Initially configured Claude 3 Sonnet, but it wasn't available in the account. Switched to Amazon Titan Text Express, which also wasn't listed. Finally found Meta Llama 3.2 available via inference profiles and got it working with the correct Llama 3 prompt format.

  2. Demo Mode vs Live API — The backend API was deployed to AWS but the frontend was running locally. Had to implement a demo mode system with hardcoded Ayurvedic responses for local development, then toggle to live Bedrock when deployed.

  3. AI Response Formatting — The Llama model initially returned long paragraphs. Had to carefully craft the system prompt to instruct bullet-point formatting with bold herb names, and build a custom markdown renderer on the frontend to convert **bold** and - bullets into proper HTML.

  4. Cycle Prediction Algorithm — Calculating accurate predictions from variable-length cycle history required averaging multiple cycles and handling edge cases like first-time users with no history.

  5. CloudFront Deployment — UTF-8 BOM encoding issues with PowerShell when creating CloudFront distribution configs. Had to use ASCII encoding workarounds for the AWS CLI JSON files.

  6. SPA Routing on S3 — React Router's client-side routing returns 403 on direct URL access. Solved by configuring CloudFront custom error responses to redirect 403 errors to index.html.

Accomplishments that we're proud of

  • Built a fully functional AI-powered Ayurvedic wellness chatbot that gives specific, actionable health advice with herb names, yoga poses, and dietary recommendations
  • Created a comprehensive period health section that normalizes menstrual conversations with personal stories, scientific information, and eco-friendly product recommendations
  • Implemented a complete cycle tracking system with mathematical prediction algorithms for next period, fertile window, and ovulation
  • Deployed a full-stack serverless application on AWS with 7 Lambda functions, 5 DynamoDB tables, Cognito auth, Bedrock AI, and CloudFront CDN
  • Designed the platform to honor Indian grandmother (Bhamma) wisdom while backing it with modern science

What we learned

  • How to integrate Amazon Bedrock with different foundation models (Claude, Titan, Llama) and handle their different API formats
  • The importance of system prompt engineering — small changes in the prompt dramatically affect response quality, formatting, and tone
  • AWS CDK for Infrastructure as Code — managing complex multi-service architectures declaratively
  • Building responsive, animated UIs with Tailwind CSS and CSS keyframe animations
  • The depth of Ayurvedic medicine — doshas, herbs like Ashwagandha, Shatavari, Brahmi, Manjistha, and their specific applications for women's health
  • Menstrual health awareness — understanding that only 18% of Indian women have access to proper hygiene, and how technology can help bridge this gap

What's next for Swasthyashe

  • Multi-language Support — Hindi, Telugu, Tamil, Kannada, Malayalam translations to reach rural India
  • AI-Powered Symptom Analysis — Use Bedrock to analyze symptom patterns over time and provide predictive health insights
  • Community Forum — A safe space for women to share experiences, ask questions, and support each other
  • Telemedicine Integration — Connect users with Ayurvedic practitioners for virtual consultations
  • Wearable Device Integration — Sync with fitness trackers for automatic cycle and health data logging
  • Personalized AI Meal Plans — Use Bedrock to generate weekly Ayurvedic meal plans based on dosha, cycle phase, and dietary preferences
  • Mobile App — React Native version for iOS and Android
  • Offline Mode — Service worker for accessing health content without internet in rural areas

Built With

Share this project:

Updates