Software Engineer · Backend · Secure Systems

I build backend systems designed to survive the real world.

I am a software engineer focused on backend engineering, reliable APIs, Linux-based systems, observability, and security-aware architecture. I treat software as something that should be maintainable, observable, predictable, and secure by design.

I teach software development at SENAI in Santana de Parnaíba, helping developers understand not only how code works, but why systems behave the way they do.

View my work Get in touch

$ whoami — software engineer, backend, secure systems, educator

Featured project

AI Code Auditor

A FastAPI service that reviews a GitHub pull request once per push—signed webhooks, deterministic checks, and a single idempotent comment.

Jump to case study

Selected work

Engineering case studies

Four public projects that show how I design APIs, integrate systems, and reason about failure—not a list of logos.

AI Code Auditor

Personal · backend service · MVP

2025

01GitHub webhook

02HMAC + HTTP 202

03Rules + LLM

04One PR comment

Problem: Pull-request review is easy to spam with noisy bot comments and hard to make consistent. The goal was a service that reviews a PR once per push, combining deterministic checks with a second-pass model review.

Approach: FastAPI accepts signed GitHub webhooks, acknowledges quickly, analyzes the unified diff with a rule engine plus an LLM adapter, scores findings, optionally persists the run, and posts a single idempotent PR comment.

Result: A testable webhook service with HMAC verification, HTTP 202 dispatch, a hidden head-SHA marker so comments update instead of stacking, and a Cloud Run path with Secret Manager. Shipped as an MVP with documented limitations.

Python
FastAPI
GitHub webhooks
Firestore
Docker
Cloud Run

Repository

Store Backend Service

Personal · REST API · cloud-shaped

2025

Auth/register · /login

CatalogProduct CRUD

OrdersJWT-protected

StoreFirestore

Problem: Build a cloud-oriented store backend with clear auth, product, and order boundaries—runnable locally with Docker and described as infrastructure, not only as application code.

Approach: FastAPI exposes register/login, product CRUD, and JWT-protected orders. Firestore persists data (emulator locally, managed in GCP). Docker Compose, Make targets, and Terraform cover health checks and a production-like run.

Result: A small backend that demonstrates REST design, authentication, data modeling, and infrastructure-as-code. The same service can be exercised locally without a live cloud account.

Python
FastAPI
JWT
Firestore
Docker
Terraform
GCP

Repository

SIEM Homelab with Elastic Stack

Personal · systems lab · observability

Lab

GenerateAttacker

EndpointUbuntu / auth.log

ShipFilebeat

IndexElasticsearch

InvestigateKibana

Problem: Security monitoring is hard to learn as disconnected tools. The lab needed a small, reproducible pipeline that shows a full event lifecycle: activity, collection, indexing, and investigation.

Approach: Docker Compose wires an Ubuntu endpoint, Filebeat, Elasticsearch, Kibana, and an attacker container used to generate authentication events such as failed SSH logins. Logs move from /var/log/auth.log into a Filebeat index that can be queried in Kibana.

Result: A minimal endpoint → forwarder → SIEM path that demonstrates systems integration, observability, Linux, and defensive security as engineering—not as a hacking showcase.

Docker Compose
Linux
Filebeat
Elasticsearch
Kibana

Repository

Extensão Santuário

Volunteer · Flutter MVP · community

2025

AuthLogin · Register

TabsHome · News · Services

DrawerHistory · Gallery · Events

CommunityRomarias · Transmissions

Problem: A community sanctuary needed a mobile surface for news, services, pilgrimages, and transmissions. The work was volunteer and unpaid: build an MVP that follows a real Figma prototype without turning the app into a cluttered menu.

Approach: Flutter implementation mapped 1:1 to the prototype. Authentication, modular feature folders, shared design tokens, and reusable widgets. Primary actions stay in a bottom bar; institutional content lives in a drawer.

Result: A navigable community MVP covering login, news, services, events, and complementary screens—showing product thinking and software structure applied outside a backend-only brief.

Flutter
Dart
Auth flows
Modular UI

Repository

About

I design backend services that stay understandable after they leave a laptop.

I like understanding systems underneath the framework. The interesting work sits where software architecture, operating systems, infrastructure, observability, and security meet.

Based in Santana de Parnaíba, I spend weekdays teaching software development and evenings in Linux labs—building APIs, watching logs, and keeping mental models honest enough to explain out loud.

The working loop is simple: Build → Observe → Break → Improve. Clear boundaries are not enough if you cannot see how a system fails. Security belongs here: it is how you design for misuse, not a separate personality.

Python
FastAPI
Flask
Java
.NET
PostgreSQL
Docker
Linux
Terraform
GitHub Actions
Elastic

Software Development Instructor at SENAI (Mar 2025–present)
Instructor in the municipal Fábrica de Programadores
Volunteer Flutter MVP for a community sanctuary app

If you are hiring for backend or systems work, message me on LinkedIn.

Social proof

What people take away from working with me

Short notes from teaching and municipal IT—how the work lands with people around it.

He does not teach syntax in isolation. He walks through why a service fails, then makes us rebuild the mental model before we write more code.

SENAI student

When something in the municipal stack was down, he treated it as a system: logs first, then the smallest change that restored the service.

Municipal IT colleague

Classroom labs stay small on purpose. The point is that a student can finish the loop and still understand every moving part.

Class collaborator

Contributed with

SENAI
Fábrica de Programadores
Prefeitura de Santana de Parnaíba
Câmara Municipal de Santana de Parnaíba

Experience

Roles

Mar 2025 — Present

Software Development Instructor

SENAI · Fábrica de Programadores

Teach programming foundations through software engineering practice: object-oriented design, data structures, and algorithms.
Guide hands-on web and backend work across JavaScript, React, Node.js, Java, datastores, version control, testing, and code review.
Cover server-side architecture and deployment habits, including CI/CD on Azure as part of the curriculum.

Mar 2023 — Mar 2025

System Administrator

Câmara Municipal de Santana de Parnaíba

Operate municipal endpoints and identity: workstations, user accounts, and access control for daily council work.
Keep legislative and administrative systems available, with backups and practical Linux/Windows administration.
Respond to incidents with a preference for documented, repeatable operational habits over one-off fixes.

Apr 2022 — Feb 2023

Data Center Analyst

Prefeitura de Santana de Parnaíba

Monitor and support municipal data-center infrastructure, with attention to service availability.
Assist with server health, networking basics, and the operational path from alert to restored service.
Hands-on ops work that later shaped a systems approach to software: observe first, then change.

Labs built with students

Classroom demos, not products

These were created with students as simple didactic classroom work. They are teaching demos—small enough to finish in class—not production systems or portfolio-grade products.

Zombie Pose Detector

A Teachable Machine pose lab used in class to show on-device inference in the browser.

Open lab

Chrome Dino Autoplayer

A PyAutoGUI and OpenCV exercise that automates Chrome’s dinosaur game to teach screen automation.

Repository

Student starter repository

A shared GitHub base students fork to organize and publish the code they write in class.

Repository

Contact

Let’s talk about a backend project.

The fastest way to reach me is LinkedIn. GitHub is where the code lives.

Message me on LinkedIn GitHub