New Year, New Tech: How I Turned 42 School Projects into a Cloud-Native App

Everton Vaz — Wed, 28 Jan 2026 21:24:39 +0000

This is a submission for the New Year, New You Portfolio Challenge Presented by Google AI

About Me

I'm a Software Engineer and 42 São Paulo graduate with a nontraditional background. After 13 years as an entrepreneur managing a fashion retail business, I pivoted to technology driven by a deep curiosity for how things work—from philosophy and physics to low-level code.

I classify myself as a "dynamic profile," blending the discipline of systems programming with the creativity of indie culture and street art. Most recently, I've been building real-time educational platforms and employability tools for NGOs, always looking to bridge the gap between complex backend logic and user value.

Portfolio

You can interact with my portfolio here:

How I Built It

This portfolio is a technical showcase connecting modern web technologies with low-level system programming.

The Tech Stack:

Frontend: React, Vite, and Tailwind CSS for a retro-futurist punk aesthetic.
Backend: Elixir (Phoenix/Bandit) acting as a high-concurrency orchestrator.
Core: Native C executables (Minishell and Dining Philosophers simulation).
Infrastructure: Docker and Google Cloud Run using multi-container deployments.

The Architecture:
Instead of just displaying static text, the application actually runs my C projects in the cloud. When you access the "Terminal" or "Philosophers" modules, the React frontend opens a WebSocket connection to the Elixir backend. Elixir then spawns the corresponding C binary as a managed process (interaction via Ports), piping stdin and stdout in real-time between the browser and the Linux container.

Google AI Tools:
I built this project paired with Google Antigravity, leveraging the advanced reasoning capabilities of Gemini Pro and Flash. The AI acted as a tireless full-stack partner, accelerating everything from the "Cyberpunk" React UI design to the complex Elixir GenServer logic required to manage C binaries via WebSockets. This collaboration allowed me to rapidly iterate on difficult technical challenges, such as real-time thread synchronization visualization and terminal emulation, bridging the gap between high-level web code and low-level system processes.

For deployment, we utilized Google Cloud Run to host the multi-container architecture. The AI assisted in configuring the robust Docker environment, ensuring that stateful C processes could run reliably and securely within the stateless container paradigm. This combination of next-generation AI agency and scalable cloud infrastructure was crucial in transforming a local system programming showcase into a globally accessible, interactive web experience.

What I'm Most Proud Of

I'm most proud of the Interoperability Layer. It is not easy to make a stateful C program like a Shell or a Thread Simulation behave well over a stateless HTTP/WebSocket connection.

Building the "BFF" (Backend for Frontend) in Elixir to supervise these C processes ensures that if a user crashes their shell or deadlocks their philosopher simulation, it doesn't affect the main server—the process just dies and restarts cleanly. This robust error isolation is a concept I learned at 42 and applied here using the Erlang VM's strengths.

DEV Community: Everton Vaz

New Year, New Tech: How I Turned 42 School Projects into a Cloud-Native App

About Me

Portfolio

How I Built It

What I'm Most Proud Of