~/gabbensa $ whoami

  role      : full-stack developer
  stack     : typescript · react · next.js · node · postgres · python
  focus     : multi-tenant saas · llm integrations · automation
  origin    : network administration (mcsa) → software
  status    : shipping something in stealth
  rule #1   : security is designed in, never bolted on

~/gabbensa $ _

Full-stack developer working across React / Next.js / TypeScript on the front, Node.js, Postgres and Supabase on the back — with a strong pull toward LLM integrations and Workflow Automation.

I started as a Network Administrator (MCSA), which is where I picked up the habit that still shapes how I build: infrastructure and security are not a later chapter. Row-Level Security, tenant isolation and auth get designed in from day one, not bolted on when something breaks.

Mostly self-driven. I learn fast, I ship, and I keep asking dumb questions until I actually understand the thing.

Languages

Front-end

Back-end & Data

AI & Automation

DevOps & Tools

My main focus right now. A multi-tenant SaaS platform, built and architected solo.

Next.js TypeScript Supabase (PostgreSQL + RLS) Railway

• Multi-tenant backend architecture with role-based access control and strict tenant data isolation
• Secure API structure, auth and Row-Level Security enforced at the database layer
• LLM-powered automation integrated into the product

Private repo — happy to walk through the architecture in a conversation.

Peer-to-peer messaging app for rescue units — communication with no cellular and no internet. Final year project, led end to end.

Java Android SDK WiFi Direct Sockets

• Automatic device discovery and socket-based communication between phones
• Real-time message sync across connected devices
• Responsive UI built with the Android SDK

Full-stack booking app — search, book and manage flights.

React TypeScript Node.js PostgreSQL GitHub Actions

• Complete front-to-back implementation
• CI/CD pipeline with automated testing and deployment

Text classification model separating AI-generated from human-written content — 99% accuracy on real datasets.

Python Naive Bayes Support Vector Classifier

Smaller builds and coursework — kept public because they show the range.

💬 Get in touch

Building something interesting? Let's talk.

bensamoungabriel@gmail.com