Skip to main content

Software engineering

Dr. Elias Dahdal Clinic — Clinic Management System

A multi-department clinic management system in production: reception, laser, dermatology and dental workflows, billing and accounting in two currencies, payroll, inventory, audit and a patient portal.

Evidence
Verified work
Category
Software engineering
Context
Chief Technology Officer (CTO) · VegaCORE
Dr. Elias Dahdal clinic system — the public sign-in page for staff and patients (Arabic, right-to-left)

Overview

The operating system of a multi-department clinic, built for the Dr. Elias Dahdal clinic and running in production: one platform for reception and scheduling, the laser, dermatology and dental departments, billing, accounting, salaries and doctor shares, inventory, notifications and an audit trail — with a separate portal for patients. The interface is Arabic, right-to-left.

Problem

A clinic with several departments runs on many people deciding at once: reception assigns specialists and rooms, each department records its own procedures, materials are consumed from a shared stock, and every payment has to be split correctly between the clinic and its doctors — in US dollars and Syrian pounds. On paper and spreadsheets, none of that can be audited.

Constraints

Each department’s records are visible only to that department and to the super admin. Operations run inside a working day that the super admin opens and closes, and the USD/SYP exchange rate is entered once at the start of each day — every price, profit and share is calculated against it. Every staff action has to be traceable, and patients need their own, strictly limited access.

Approach

The work started from a written requirements specification agreed with the clinic — roles, departments, procedure catalogues, pricing and the financial rules — and was built as a role-based system in which every role works in its own workspace and sees only what it is allowed to.

Architecture

A React + TypeScript front end (Vite) and a Node.js/Express REST API over MongoDB (Mongoose). The API is organised into 23 route modules — patients, scheduling, rooms, laser, dermatology, skin, dental, solarium, clinical sessions, billing, accounting, finance, salaries, doctor shares, inventory, notifications, reports, audit, users and the patient portal — with JWT authentication kept separate for staff and patients.

Deployment is automated: a GitHub Actions workflow deploys to the VPS over SSH and restarts the API under pm2.

Engineering decisions

A “business day” record gates daily operations and stores that day’s exchange rate, so financial records cannot drift from the rate they were created under.

Authorisation follows the clinic’s organisation: eight staff roles (super admin, reception, laser, dermatology staff, manager and assistant manager, dental branch and dental assistant) map onto route-level access rules.

Dermatology sessions draw their materials from the clinic inventory and report items that are running low; doctor shares are computed from percentages the super admin configures.

Results

In production at the clinic (eliasdahdal.clinic). The patient portal covers appointments, records, a financial statement, profile and account security, and opens with a 3D welcome scene of the clinic built with React Three Fiber.

Software engineering

Tavola — Multi-Tenant Restaurant Reservation Platform

A multi-tenant SaaS for restaurant reservations, tables, branches and staff: NestJS backend with real-time updates, a Flutter customer app, a React restaurant dashboard and a platform-owner console.

Stack

  • NestJS
  • TypeScript
  • PostgreSQL
  • Prisma
  • Redis
  • BullMQ
  • Real-time systems (WebSockets)
  • Multi-tenant SaaS architecture
  • Authentication & session security
  • Authorization (RBAC)
  • Object storage (S3 / MinIO)
  • Docker
  • Nginx
  • Flutter
  • Dart
  • React
  • Tailwind CSS
  • Internationalization & RTL
  • System design & ADRs
  • Automated testing
  • Source
  • Documented: Problem, Approach, Architecture, Results

Artificial intelligence

AI Intelligence Project Management System

AI-powered enterprise project and task management platform with intelligent task generation and smart assignment.

Stack

  • Large language models
  • Neural networks
  • FAISS vector retrieval
  • Expert systems
  • Algorithm design
  • Node.js
  • Express.js
  • FastAPI
  • MongoDB
  • Next.js
  • React
  • Flutter
  • Authorization (RBAC)
  • Python
  • SQL database engineering
  • 5 figures
  • Documented: Problem, Approach, Architecture, Intelligence layer, Results

Software engineering

VegaCore Operating System (VOS)

An ERP / CRM / project-management platform for VegaCORE: CRM, projects, marketing, media, HR, finance, archive and AI modules on NestJS, Prisma, PostgreSQL and Next.js 15.

Stack

  • NestJS
  • TypeScript
  • PostgreSQL
  • Prisma
  • Redis
  • Object storage (S3 / MinIO)
  • Next.js
  • Tailwind CSS
  • Authorization (RBAC)
  • Authentication & session security
  • Docker
  • Nginx
  • Source
  • Documented: Architecture