FAP Next Generation

AI-enabled digital platform for the Family Adoption Programme (FAP) under NMC-CBME

Live Demo Links

• Live application: https://fap-nextgen-app.vercel.app
• Summit demo video: https://drive.google.com/file/d/1bUuuoM_xp_lIhGCuo_-dW5zvyExifNL2/view

Why This Project Exists

Family Adoption Programme implementation in many institutions is still paper-heavy, fragmented, and difficult to audit at scale.
FAP Next Generation was built to convert mandatory fieldwork into:

• better student learning quality
• stronger mentor oversight
• cleaner institution-level evidence
• actionable community health intelligence

Vision Statement

Move from paper logbooks to a longitudinal, AI-assisted learning and public-health intelligence system for community medicine training in India.

Journey: From Start to Current State

Stage	Focus	Outcome
Phase 1	Digital family records and field logbook workflows	End-to-end student capture of family, visit, and assessment data
Phase 2	Role-based governance and mentor workflows	Student, Teacher, Admin pathways with review and grading loops
Phase 3	AI integration for reflection quality	Gibbs-cycle extraction, quality flags, safety controls, confidence metadata
Phase 4	Programmatic scale readiness	Multi-provider AI keys, fallback controls, offline-first sync, exportable reports

Platform At a Glance

flowchart LR
    A[Student Field Visit] --> B[Family + Member + Assessment Capture]
    B --> C[Reflection Upload or Structured Entry]
    C --> D[AI Extraction + Gibbs Segmentation]
    D --> E[Mentor Review + Rubric Grading]
    E --> F[Reports + Logbook Exports]
    B --> G[Community Analytics]
    G --> H[Institution Planning Insights]

Loading

Architecture

flowchart TB
    UI[React + Vite PWA] --> AUTH[Supabase Auth]
    UI --> DB[(Supabase Postgres + RLS)]
    UI --> STORAGE[Supabase Storage]
    UI --> CACHE[IndexedDB + React Query Persist]
    UI --> AI[Multi-Provider AI Layer]
    AI --> OR[OpenRouter]
    AI --> GOOG[Google AI Studio]
    AI --> OA[OpenAI]
    AI --> MISC[Other configured providers]
    UI -. optional .-> MICRO[Micro AI Service /v1 ingest job result]

Loading

Core Modules

• Student Dashboard
  • adopted families, population, issues, activity summary
• Family Folder
  • family + member records, longitudinal continuity, visit logs
• Structured Assessments
  • NCD, socio-economic, nutrition, mental health, maternal-child focused forms
• Reflections
  • structured entry or upload, AI segmentation into Gibbs stages
• Learning Objectives
  • competency-linked objectives and expected activities
• AI Medical Coach
  • context-aware guidance for community medicine scenarios
• Community Health Profile
  • overview, demography, resources, annual planning, health status
• Mentor Workspace
  • pending reviews, AI-segmented entries, rubric-based assessment
• Admin Portal
  • user management, role governance, provider/system settings
• Reports & Logbook
  • print/export workflows and summary analytics

AI Core: Gibbs Reflective Learning Engine

The reflection engine is designed for formative learning quality, not just text generation.

• input modes:
  • structured typing
  • uploaded file extraction (doc/pdf/image pipeline)
• stage mapping:
  • Description
  • Feelings
  • Evaluation
  • Analysis
  • Conclusion
  • Action Plan
• quality intelligence:
  • missing-stage flags
  • section-length checks
  • confidence scores
  • evidence spans
  • safety disclaimers and diagnosis-claim guardrails
• reliability controls:
  • provider fallback policies
  • micro-AI pipeline preference toggle
  • AI audit/version tables for traceability

Feature Matrix

Area	Capability	Benefit
Data Capture	Family, member, visit, assessment logging	Continuity of care and complete field documentation
AI Reflection	Gibbs segmentation + quality checks	Better reflective depth and clinical reasoning
Mentor Tools	Pending triage + rubric scoring	Higher feedback quality with less administrative load
Community View	Population and health indicator summaries	Program-level planning and public health insight
Reports	Exportable logbook and summaries	Compliance-ready outputs for reviews and assessment
Operations	Role-based access + RLS	Privacy, governance, and secure scale
Connectivity	Offline-first patterns + sync	Reliable use in low-connectivity environments

National Alignment

National Priority / Framework	Alignment in FAP Next Generation
NMC CBME (UGME 2023 context)	Structured competency-linked workflows and longitudinal field documentation
Family Adoption Programme (FAP)	Purpose-built digital execution of family-level community immersion
Public health priority programs	NCD, maternal-child, nutrition, TB-oriented community workflows and resources
Ayushman Bharat direction	Strengthens primary-care community data practices at grassroots level
ABDM-ready ecosystem thinking	ABHA-linked fields and interoperability-oriented architecture path
Digital health governance	Role-based controls, auditability, institution-level manageability

Benefits by Stakeholder

Students

• stepwise reflective learning support
• easier field documentation and continuity
• better preparation for competency-based evaluations

Mentors / Faculty

• targeted review on high-priority submissions
• rubric consistency across cohorts
• reduced manual burden

Institutions

• auditable, exportable records
• stronger quality monitoring
• better evidence for academic and administrative review

Public Health Systems

• structured community-level signals from routine educational fieldwork
• potential for improved local planning and early trend visibility

Technology Stack

Layer	Stack
Frontend	React, Vite, React Router
Offline & Cache	IndexedDB, React Query persistence, PWA
Backend	Supabase (PostgreSQL, Auth, Storage, RLS)
AI Layer	Multi-provider key architecture (OpenRouter, Google, OpenAI, and others)
Optional AI Microservice	FastAPI + async job endpoints (/v1/ingest, /v1/job/{id}, /v1/result/{id})
Deployment	Vercel

Quick Start

1) Prerequisites

• Node.js 18+
• npm
• Supabase project
• at least one AI provider API key

2) Clone and install

git clone https://github.com/hssling/FAP_Nextgen_App.git
cd FAP_Nextgen_App
npm install

3) Configure environment

Create .env from .env.example and set the values:

VITE_SUPABASE_URL=...
VITE_SUPABASE_ANON_KEY=...
VITE_OPENROUTER_API_KEY=...
VITE_GOOGLE_AI_KEY=...
VITE_OPENAI_API_KEY=...
VITE_MISTRAL_API_KEY=...
VITE_XAI_API_KEY=...   # or VITE_xAI_API_KEY
VITE_CEREBRAS_API_KEY=...
VITE_HUGGINGFACE_API_KEY=...
VITE_MICRO_AI_BASE_URL=http://localhost:8000

4) Run

npm run dev

5) Build

npm run build
npm run preview

Security, Privacy, and Ethics

• Supabase Auth + role-based route protection
• Row Level Security policies for data isolation
• local key management options for AI providers
• safety boundary in AI outputs:
  • decision-support framing
  • anti-definitive-diagnosis guardrails

Repository Structure

src/
  components/
  contexts/
  data/
  pages/
  services/
  utils/
supabase/
micro_ai_service/

Roadmap Potential

• multilingual UI and reflection support (regional language-first workflows)
• deeper longitudinal analytics and cohort benchmarking
• stronger PHC/CHC decision dashboards
• interoperability expansion for broader digital health ecosystems
• institution-scale onboarding automation

Documentation Archive

• Previous README archived as: README_ARCHIVE_2026-02-15.md

Acknowledgement

Developed as a community medicine and medical education innovation aligned to the goals of competency-based training and digitally enabled public health practice.