SymbioticaAI

Inspiration

In India, over 10,000 children are born with Thalassemia every year, many without access to regular blood transfusions or adequate medical care. Through researching about the Blood Warriors initiative, I was deeply moved by the logistical challenges patients face — from locating timely donors to managing treatment plans, often with minimal support. As a final year Computer Science student passionate about healthcare and AI, I saw an opportunity to build a technology-driven system that doesn't just react, but anticipates and supports — bridging the gap between patients, donors, and healthcare providers. That’s how SymbioticaAI was born — an intelligent, multi-agent platform focused on predictive blood donation, personalized care, and real-time communication for Thalassemia patients.

What it does

SymbioticaAI is an AI-powered, cloud based platform that: Predicts blood donor availability using historical data and behavior modeling. Matches Thalassemia patients with compatible donors in real-time, based on urgency, location, and blood group. Provides a multilingual conversational AI assistant and support patients with personalized care guidance. Summarizes medical reports using NLP for doctor-patient clarity. Tracks and gamifies donor contributions to drive recurring participation. Offers a centralized dashboard for NGOs and medical staff to monitor transfusions, donor engagement, and patient well-being.

How we built it

We designed SymbioticaAI as a modular, scalable platform using a mix of AI, NLP, cloud services, and frontend technologies: Architecture: Frontend: React.js for web dashboard, with responsive views for patients and donors. Backend: FastAPI on AWS Lambda for serverless architecture. AI/ML:Donor availability prediction using Scikit-learn time series models. HuggingFace Transformers (BERT) for intent classification and patient need detection. Amazon Comprehend Medical for entity extraction from medical texts (diagnosis, meds, labs). Amazon Textract for parsing handwritten lab reports and scanned PDFs. Chatbot: Built using Amazon Lex, integrated with AWS Translate for language support. Data: DynamoDB for real-time donor-patient mapping. PostgreSQL (RDS) for structured patient records. Notifications: AWS SNS + WhatsApp Cloud API for alerts and reminders. Security: AWS Cognito for authentication + encryption for secure health data handling.

Challenges we ran into

Lack of public donor data: We simulated realistic datasets using Blood Warriors’ published insights and research papers. Data privacy compliance: Ensuring patient and donor data remained confidential required careful design with encrypted databases and strict role-based access control. Real-time matching logistics: Balancing urgency, compatibility, and donor fatigue was a challenge for algorithm design under time constraints.

Accomplishments that we're proud of

Successfully implemented a donor matching system with predictive modeling and intelligent fallback handling. Created a multilingual chatbot capable of basic care guidance for patients in regional languages. Enabled document summarization of lab reports using NLP — making medical information more accessible for families. Designed a system that’s not only technically scalable, but socially impactful and immediately applicable in the real world.

What we learned

How to architect a serverless AI system on AWS that is cost-effective and scalable. Practical applications of NLP in healthcare, especially for underserved populations. The importance of UX in health-tech — especially when designing for patients, donors, and NGOs with varying tech exposure.

What's next for SymbioticaAI

Expanding donor reach through integration with e-RaktKosh and other national blood bank systems. Adding more intelligent health insights, such as hemoglobin tracking. Launching a mobile-first app for rural accessibility and offline data sync. Partnering with government or CSR programs to deploy in high-need regions.

Built With

• amazon-comprehend-medical
• amazon-lex
• amazon-textract
• amazon-translate
• api-gateway
• aws-cognito
• aws-lambda
• aws-sns
• cloudwatch
• dynamodb
• ec2
• fastapi
• git
• github
• huggingface-transformers-(bert)
• iam
• oauth-2.0
• python-3.10
• rds-(postgresql)
• react-native
• react.js
• s3
• scikit-learn
• spacy
• tailwind-css
• twilio
• whatsapp-cloud-api