AI-Powered Unified Clinic Notes & Billing System

📌 Project Overview

The AI-Powered Unified Clinic Notes & Billing System is a full-stack application designed to simplify clinic workflows by allowing doctors to enter prescriptions, lab tests, and clinical notes in a single unified interface. The system uses AI to automatically classify unstructured medical text into structured data and generates billing and printable PDF reports.

This project demonstrates practical integration of AI/NLP with a scalable software architecture, focusing on usability, structured data storage, billing accuracy, and report generation.

---

🚀 Features

• Unified clinic input screen for doctors

• AI-powered medical text classification

• Automatic extraction of:

  • Drugs & dosages
  • Lab tests
  • Clinical notes

• Structured database storage

• Billing calculation based on drugs and lab tests

• Printable PDF medical report

• Clean REST API architecture

• Modular backend design

• Responsive NextJS 13 frontend

---

🛠️ Tech Stack

Backend

• Golang
• Echo Framework
• PostgreSQL
• AI API (LLM-based text classification)
• gofpdf (PDF generation)

Frontend

• NextJS 13
• TypeScript
• React Query
• Axios

Database

• PostgreSQL

Tools

• Docker (optional)
• Postman
• Draw.io (Architecture diagrams)

---

🏗️ System Architecture

The system follows a layered architecture with clear separation of concerns.

React Frontend
      ↓
Golang API (Echo)
      ↓
Service Layer
      ↓
AI Service
      ↓
Repository Layer
      ↓
PostgreSQL Database

External Services

• AI API for NLP classification
• PDF generator for report printing

---

🧠 AI Integration

AI is used to convert unstructured medical text into structured clinical data.

Workflow

1. Doctor enters medical text

2. Backend sends text to AI service

3. AI extracts:

   • Drugs
   • Lab tests
   • Clinical notes

4. Data is validated

5. Structured data stored in PostgreSQL

6. Billing calculated

7. PDF report generated

Example

Input

Patient has fever for 3 days. Prescribe Paracetamol 500mg twice daily. Order CBC test.

AI Output

Drugs:
Paracetamol 500mg

Lab Tests:
CBC

Notes:
Patient has fever for 3 days

---

🗄️ Database Design

The database is designed using normalized relational tables.

Main Tables

• patients
• visits
• prescribed_drugs
• lab_tests
• clinical_notes
• drug_catalog
• lab_test_catalog
• billing

Design Approach

• Catalog tables store official prices
• Visit tables store structured AI output
• Snapshot pricing ensures billing consistency
• Raw AI input is stored for auditing

---

💰 Billing Logic

Billing is calculated using:

Grand Total = Consultation Fee + Drug Prices + Lab Test Prices

Pricing Strategy

• Drug prices fetched from drug_catalog
• Lab test prices fetched from lab_test_catalog
• Prices stored in visit tables
• Billing stored in billing table

Benefits

• Consistent billing
• Historical price tracking
• Easy invoice generation

---

🖨️ PDF Report Generation

The system generates a printable PDF report for each visit.

Report Includes

• Clinic information
• Patient details
• Prescription
• Lab tests
• Clinical notes
• Billing summary

API

GET /visits/{id}/report

Output

PDF file ready for printing or download.

---

🌐 API Endpoints

Visits

POST /visits
Create new visit and process AI classification

GET /visits
Get all visits

GET /visits/{id}
Get visit details

Billing

GET /billing/{visit_id}
Get billing details

Report

GET /visits/{id}/report
Download PDF invoice

Generate new documentation for api endpoint

swag init -g main.go -d cmd/service,internal

---

🖥️ Frontend Workflow

1. Doctor enters medical text
2. AI processes text
3. Structured data displayed
4. Billing calculated
5. PDF invoice generated
6. User downloads invoice

---

⚙️ Setup Instructions

Prerequisites

• Go 1.25+
• Node.js 18+
• PostgreSQL
• AI API Key

---

🔧 Backend Setup

1. Clone Repository

git clone https://github.com/KNR1997/21c_Care.git
cd 21c_Care

2. Configure Environment

Create .env

PORT=8080
DB_HOST=localhost
DB_PORT=5432
DB_USER=postgres
DB_PASSWORD=postgres
DB_NAME=clinic
AI_API_KEY=your_api_key

3. Run Database

psql -U postgres
CREATE DATABASE clinic;

4. Run Migrations (GOOSE command)

goose -dir ./migrations postgres "host=localhost user=myuser password=mypassword dbname=mydatabase port=5432 sslmode=disable" up

5. Start Server

air

Server runs on:

http://localhost:7788

---

🎨 Frontend Setup

1. Go to frontend

cd web

2. Install dependencies

yarn

3. Configure environment

copy `.env.template` to `.env`

4. Run

yarn dev

App runs on:

http://localhost:3002

---

🧪 Example Workflow

Step 1

Doctor enters:

Patient has headache. Prescribe Ibuprofen 200mg. Order MRI.

Step 2

AI extracts:

• Ibuprofen
• MRI
• headache

Step 3

Database stores structured data

Step 4

Billing generated

Step 5

PDF report downloaded

---

📊 Project Structure

backend/
    internal/
        handlers/
        services/
        repositories/
        ai/
        report/
    database/
    main.go

frontend/
    src/
        components/
        pages/
        data/
        client/

---

🧠 Assumptions

• AI classification may not be 100% accurate
• Drug and lab test catalog contains common items
• Consultation fee is fixed
• Single clinic environment
• Internet connection required for AI API

---

👨‍💻 Author

Software Engineer Candidate

AI-Powered Clinic System Technical Assessment