Swasthya AI is an AI-powered preventive healthcare platform designed to shift healthcare from reactive treatment to proactive monitoring.
It enables continuous health tracking through a memory-aware conversational AI, real-time risk analysis, and intelligent doctor support systems — ensuring early detection, improved outcomes, and safer medical decisions.
Swasthya AI builds a continuous health intelligence loop where:
- Patients interact daily with an AI assistant
- The system learns and tracks their health patterns over time
- Risk scores are dynamically generated
- Doctors are automatically notified when intervention is needed
This creates a data-driven, proactive healthcare ecosystem instead of episodic doctor visits.
- 🧠 Memory-aware AI conversations
- 📊 Continuous health monitoring
⚠️ Early risk detection- 👨⚕️ Intelligent doctor insights
- 💊 Medication safety checks
- 👨👩👧👦 Family-level health intelligence
- Natural chat-based interaction
- Understands patient history and context
- References past symptoms intelligently
- Generates personalized follow-up questions
-
AI-generated personalized questions
-
Based on:
- Past symptoms
- Medical conditions
- Behavioral patterns
-
Captures daily health signals
-
Converts natural language into structured medical data
-
Extracts:
- Symptom type
- Severity
- Duration
- Body location
-
Enables downstream analytics and tracking
-
Combines multiple inputs:
- Symptoms
- Chronic conditions
- Medication adherence
- Family history
- Behavioral signals
-
Outputs:
- Risk score (0–100)
- Risk category
- Clinical reasoning
- Confidence level
- Visual representation of health risks across body zones
- Color-coded severity indicators
- Real-time updates based on new data
-
Detects potential drug interactions
-
Provides:
- Severity classification
- Plain-language explanation
- Safety recommendations
- Monitors medication intake patterns
- Detects missed doses
- Generates alerts for critical non-adherence
- Adapts reminder timing based on behavior
- Aggregates health trends across family members
- Identifies shared symptom patterns
- Detects potential environmental or infectious signals
- Maintains strict individual data privacy
-
Matches patients with relevant healthcare schemes
-
Based on:
- Condition
- Age
- Socio-economic profile
-
Provides eligibility and benefit details
-
Prioritized list of patients needing attention
-
Generated based on:
- Risk levels
- Alerts
- Behavioral anomalies
- Longitudinal health data view
- Risk trends and event markers
- Symptom timelines
- Adherence insights
-
Consolidated patient overview
-
Includes:
- Symptom progression
- Risk evolution
- Behavioral patterns
- Doctors can query patient data in plain language
- Responses are generated strictly from available data
- Includes source references for traceability
-
If required data is missing:
- System generates a question
- Delivered to patient in next interaction
- Response is captured and relayed back to doctor
Orchestration Layer → n8n
Processing Layer → FastAPI
AI Layer → LLM (Groq)
Data Layer → Supabase (PostgreSQL)
- Patient sends message
- Orchestration layer triggers backend
- AI processes input and extracts structured data
- Data stored in database
- Session summary generated
- Risk score computed
- Alerts triggered if thresholds are exceeded
- Doctor dashboard updated
- Context-aware conversation generation
- Structured JSON output enforcement
- Symptom extraction from natural language
- Clinical summarization
- Data-grounded QnA system
-
Rule-based scoring using:
- Conditions
- Symptoms
- Adherence
- Demographics
- Adjusts score within a bounded range
- Uses contextual reasoning
-
Based on:
- Data availability
- Interaction history
- Profile completeness
-
Uses structured knowledge sources
-
Retrieves relevant medical guidelines
-
Enhances:
- Risk explanations
- Clinical summaries
-
Ensures explainability and grounding
- Chat Processing Service
- Symptom Extraction Service
- Risk Calculation Engine
- Doctor Query Engine
- Safety Monitoring Service
- Scheme Matching Engine
- Session-based triggers
- Scheduled monitoring jobs
- Real-time alert generation
- Asynchronous data collection loops
- Python (FastAPI)
- Pydantic (data validation)
- httpx (API communication)
- n8n (workflow automation)
- Groq (LLaMA models)
- LangChain (RAG pipeline)
- Sentence Transformers (embeddings)
- Supabase (PostgreSQL)
- Real-time data sync
- React Native (patient interface)
- Next.js (doctor dashboard)
- SVG / Three.js (body heatmap)
- OpenFDA (drug interaction data)
- Firebase (authentication & notifications)
- Wearable APIs (Google Fit, Apple Health)
- Schema validation for all AI outputs
- No unverified data stored
- Deterministic decision layers for critical logic
- Strict grounding for doctor-facing responses
- Graceful fallback mechanisms
- Advanced predictive analytics
- Multi-language support
- Real-time wearable integration
- Personalized health recommendations
- Hospital system integrations
Swasthya AI transforms healthcare into a continuous, intelligent, and proactive system, enabling better decisions, early interventions, and improved patient outcomes through the integration of AI, data, and automation.