CuraLink — Patient & Researcher Discovery Platform (MVP)

CuraLink is an AI-assisted platform designed to connect patients and researchers by simplifying discovery of clinical trials, health experts, and medical publications.

This MVP was built for a healthcare hackathon challenge, focusing on clean UI/UX, personalized recommendations, and a fully functional dual-user flow (Patients + Researchers) using Next.js 14 App Router.

---

Professional Git Workflow (Used in This Project)

Throughout the CuraLink MVP development, I followed a clean, industry-standard Git workflow. Each feature was isolated, documented, reviewed, and merged properly.

Branch List Snapshot

feat/add-trial-page feat/onboarding-upgrade feat/patient-dashboard-ui feat/patient-experts-page feat/patient-favorites-page feat/patient-forums-page feat/patient-publications-page feat/patient-trials-page feat/researcher-collaborators feat/researcher-dashboard feat/researcher-favorites feat/researcher-forums fix/patient-dashboard-links main refractor/landing-page refractor/patient-researcher-sidebar refractor/sidebar

🚀 Features Overview

✅ Patient Experience

Feature	Status
Patient onboarding (condition, location, tags)	✅ Completed
Personalized dashboard	✅ Completed
Clinical trials search + status filter	✅ Completed
Health experts search	✅ Completed
Publications search	✅ Completed
AI-generated summaries (placeholder)	✅ Completed
Save to Favorites (trials, experts, publications)	✅ Completed
Patient forums (ask + view threads)	✅ Completed

✅ Researcher Experience

Feature	Status
Researcher dashboard	✅ Completed
Add New Trial form	✅ Completed
Collaborator search	✅ Completed
Save collaborators to Favorites	✅ Completed
Respond to patient questions in forums	✅ Completed

✅ Shared System

• LocalStorage-based Favorites system
• Dynamic routing for forum threads
• Clean SaaS-style sidebar layouts
• Role-based layout separation (Patient vs Researcher)
• Centralized JSON mock data for MVP
• Premium design and optimized UI spacing

---

📸 Screenshots

You can paste real app screenshots here: Landing-Page Patient-Dashboard Patient-Onboarding researcher-Onboarding

---

🧩 Project Structure (Next.js App Router)

app ├── components │   ├── PatientSidebar.tsx │   └── ResearcherSidebar.tsx ├── data │   ├── collaborators.json │   ├── experts.json │   ├── forums.json │   ├── publications.json │   ├── researcher-forums.json │   ├── researcher-trials.json │   └── trials.json ├── favicon.ico ├── globals.css ├── hooks │   └── useFavorites.ts ├── layout.tsx ├── page.tsx ├── patient │   ├── clinical-trials │   │   └── page.tsx │   ├── dashboard │   │   └── page.tsx │   ├── experts │   │   └── page.tsx │   ├── favorites │   │   └── page.tsx │   ├── forums │   │   ├── ask │   │   ├── page.tsx │   │   └── [postId] │   ├── layout.tsx │   ├── onboarding │   │   └── page.tsx │   └── publications │   └── page.tsx └── researcher ├── add-trial │   └── page.tsx ├── collaborators │   └── page.tsx ├── dashboard │   └── page.tsx ├── favorites │   └── page.tsx ├── forums │   ├── [id] │   └── page.tsx ├── layout.tsx └── onboarding └── page.tsx

---

🧪 Tech Stack

• Next.js 14 (App Router)
• React 18
• Tailwind CSS
• React Icons
• LocalStorage (for favorites + profile onboarding)
• Mock JSON data for trials, experts, publications, forums

No backend is required for this MVP.

---

🧠 Personalization Logic

When a patient completes onboarding, a profile is saved:

{
  "condition": "Lung Cancer",
  "location": "India",
  "tags": ["Tumor"]
}

All patient pages use this data to provide auto-filtered, personalized results:

Clinical trials → match trial.condition

Experts → match expert.specialty / condition

Publications → match tags (if added)

This gives a realistic “AI-powered recommendation” feel.

⭐ Favorites System

The Favorites system uses a universal React hook: useFavorites()

✅ Save ✅ Remove ✅ Check if saved ✅ Persistent with localStorage ✅ Unique key system (trial-1, expert-3, etc.)

Supports:

Trials

Experts

Publications

Collaborators

📚 Forums System

Patient:

Ask questions

View question threads

Researcher:

View patient questions

Respond with answers

Each question resolves via dynamic route: /patient/forums/[postId] /researcher/forums/[id]

🛠️ Setup & Installation

1. Clone the repo

git clone https://github.com/Khan4218/curalink-mvp.git cd curalink

2. Install dependencies

npm install

or

yarn install

3. Run the app

npm run dev

App will be available at: 👉 http://localhost:3000

📦 Deployment (Vercel)

Deployed on vercel: https://curalink-mvp.vercel.app/

🎥 Demo Video (MVP)

https://loom.com/share/folder/58055b3816de486e8316eb156d9b72b3

🎥 Backend Architecture Overview

https://www.loom.com/share/e9857848263140fe8ee1377c8e1adf46

👤 Author

Akbar Khan Full Stack Developer Portfolio: https://mir-akbar-portfolio-react-v2.netlify.app/

GitHub: https://github.com/Khan4218

This MVP meets all requirements for the hackathon challenge:

✅ Dual-flow: Patients & Researchers ✅ Personalized dashboards ✅ Trial/expert/publication discovery ✅ Forums ✅ Favorites ✅ Clean SaaS UI ✅ Production-ready structure

If you're reviewing this project: Thank you — I hope you enjoy exploring CuraLink!

🏗️ Backend Architecture (Planned)

Although this MVP uses mock JSON data for speed (as recommended for the hackathon), the full production version of CuraLink will be powered by a scalable and secure backend architecture built with Node.js / Express or FastAPI, backed by PostgreSQL.

This section outlines how the backend will be implemented when the project moves beyond MVP.

1. Tech Stack (Backend)

Layer	Technology	Purpose
Backend Framework	Node.js (Express) or FastAPI	REST API
Database	PostgreSQL	Complex relational data (trials, experts, users)
ORM	Prisma / SQLAlchemy	Schema modeling + migrations
Authentication	JWT / OAuth	Secure access control
AI Layer	OpenAI API / LLMs	Generating trial + publication summaries
External APIs	PubMed, ORCID, ResearchGate SCRAPE, ClinicalTrials.gov	Live data syncing
Deployment	Vercel (Frontend) + Render/Fly.io (Backend)	Hosted environment

2. High-Level Architecture Diagram

                ┌───────────────────────────┐
                │         Frontend           │
                │      Next.js (React)       │
                └──────────────┬─────────────┘
                               │
                    HTTPS REST API Calls
                               │
                ┌──────────────┴─────────────┐
                │         Backend API         │
                │ Node.js / FastAPI Server    │
                ├──────────────┬─────────────┤
                │   Auth       │  AI Summaries│
                ├──────────────┼─────────────┤
                │ External API Integrations   │
                └──────────────┬─────────────┘
                               │
                       SQL Queries
                               │
                ┌──────────────┴─────────────┐
                │         PostgreSQL          │
                │  Users / Experts / Trials   │
                │ Publications / Favorites    │
                └────────────────────────────┘

Database Schema (Initial)

Users Table

Column	Type	Notes
id	UUID	Primary key
role	ENUM('patient','researcher')	User type
name	TEXT	Optional
email	TEXT	Auth
condition	TEXT	For personalized results
location	TEXT	City/Country
created_at	TIMESTAMP	Default now()

Clinical Trials Table

Column	Type
id	UUID
title	TEXT
condition	TEXT
status	TEXT
investigator_id	UUID (FK researchers.id)
summary	TEXT
location	TEXT
contact	TEXT

Experts Table

Column	Type
id	UUID
name	TEXT
specialty	TEXT
condition	TEXT
experience	TEXT
available	BOOLEAN
contact	TEXT

Favorites Table

Column	Type
id	UUID
user_id	UUID (FK users.id)
item_type	ENUM('trial','expert','publication')
item_id	UUID
created_at	TIMESTAMP

4. Planned API Endpoints (REST)

Authentication POST /auth/register POST /auth/login GET /auth/me

Patient Personalization GET /patients/:id/profile PUT /patients/:id/profile

Clinical Trials GET /trials GET /trials/:id POST /trials (researcher only) PUT /trials/:id

Experts GET /experts GET /experts/:id

Publications GET /publications GET /publications/:id

Forums GET /forums POST /forums GET /forums/:id POST /forums/:id/answer (researcher only)

Favorites GET /users/:id/favorites POST /favorites DELETE /favorites/:item_id

External API Integrations

ClinicalTrials.gov

Endpoint: https://clinicaltrials.gov/api/v2/studies

Used for: pulling real trial data using condition keywords

PubMed

Used for: fetching live publications related to conditions

ORCID

Used for: researcher profile auto-imports

AI Summaries

Powered by OpenAI GPT models

Summary generated on:

Clinical Trials

Publications

Researcher profiles

Backend Deployment Plan

Backend Hosting

Render (free, simple) or Fly.io (global edge hosting)

Database

PostgreSQL on Neon.tech or Supabase

CI/CD

GitHub Actions → auto deploy to Render

Separate environments: dev, prod