Projects

I’ve been an active contributor to the Debian project since 2020. My involvement includes:

• Gave my first talk at MiniDebConf Online Brazil 2020.
• Became a Debian Maintainer, 40+ changelog entries and contributions to projects like debci and image-finder.
• Reviewed multiple package uploads and mentored newcomers.
• Presented at DebConf22 in Kosovo.
• Helped organize MiniDebConf Brasília 2023.
• Created educational videos on PeerTube.
• Wrote documentation on the Debian Brasil Wiki.
• Presented at DebConf23 in India.
• Helped organize MiniDebConf Brazil 2024 in Belo Horizonte.
• Helped organize MiniDebConf Brazil 2025 in Maceió.
• Helped launch a Debian packaging module as part of a university course, mentoring ~15 students per semester.
• Became Debian Developer in 2025.

• Kernel Team: Although I’m not a member of the kernel team, I collaborate with them from time to time, primarily by supporting Debian packaging efforts.

• Load Balancer as a Service (LBaaS): I’ve been working on this team for almost 2 years, developing a LBaaS product. It’s a multi-tier load balancing solution, designed with a Kubernetes-like architecture consisting of a control plane and multiple agents. My work spans several layers of the system, from infrastructure to service logic, using tools and technologies including Golang, Incus, Envoy, IPVS, Ansible, OpenTofu, Openstack, OVN, Prometheus, Grafana, OpenSearch, Netplan, NFtables.

• AutoSeg: I worked on this team for 2 years and 4 months, maintaining an Identity and Access Management (IAM) system. My responsibilities included:

  • Setting up the IAM for new clients, performing QA testing, validating production deployments, and providing post-deploy support.
  • Communicating directly with customers to troubleshoot issues, offer support, and assist in decision-making. I also wrote user stories based on their business requirements.
  • Developing new features for the IAM, primarily integrations with third-party APIs.
  • The application was built with Ruby on Rails, with additional scripts in Bash and Python.

I was part of the Infra/DevOps team, where I worked on improving deployment processes, monitoring infrastructure, and managing virtualized environments. My work included:

• Implementing continuous deployment pipelines using GitLab CI, Docker, and Rancher API.
• Creating and maintaining GitLab CI pipelines for all microservices, including a reusable pipeline structure to simplify development workflows.
• Configuring and managing Nginx as both a web server and reverse proxy.
• Monitoring infrastructure with M/Monit and using the Elasticsearch stack for container observability.
• Writing and maintaining numerous Dockerfiles and docker-compose configurations.
• Managing virtual machines using virsh, and creating VM images with QEMU.

• The use of the acceleration strategy in derivative-free methods for solving nonlinear systems: solving large-scale nonlinear systems without derivatives, employing the spectral residual method to reduce computational costs. It evaluates the effectiveness of a new secant acceleration technique, by integrating it into an existing R implementation and comparing it to the unaccelerated version. Here is the implementation.

• Dependencies manager: a planner that solves the package installation problem. The problem was simplified in order to treat all dependency rules equally and disregard the package versions.

• Pseudo-Boolean approach to virtual machine consolidation: The problem consists in placing a set of virtual machines in a set of hardware in a way to increase workload on hardware where they can operate more energy-efficient, in this project we proposed a formulation based on this article. We extended the original approach by introducing a new overbooking scenario.

• ESP32: IoT project to play with ESP-IDF, ThingsBoard, NVS (Non-Volatile Storage), MQTT protocol, Wifi and GPIO.

• Air Fryer: a system that simulates the control of an Air Fryer with digital control.

• Semaphore Control System: a distributed system for controlling and monitoring traffic signals using Raspberry Pi boards. A central server handled control and the user interface, while distributed nodes managed local signal control and sensor monitoring, including pedestrian buttons, vehicle sensors, speed tracking, and red light violations.

• Ta na Mesa: a web application to manage restaurant orders.

• Regularization methods for nonlinear programming: this research explored the adaptive regularization method of order p+1 for nonlinear programming and validated it against classical approaches.

• Topics in linear programming and computational linear algebra: this research explored the conjugate gradient method as a solver for a subproblem in Newton's method for unconstrained nonlinear systems.

• ArBC: a web application based on the AR (Augmented Reality) APIs AR.js and A-Frame, which aimed to make the literacy process more interactive and engaging.

• Battleship game: my first project involving multiple files and object orientation. I remember having a lot of fun at the time writing C++.