Luiz Filho

At Weaveworks, I actively contributed to various open-source projects, creating and maintaining multiple Kubernetes controllers. This hands-on experience deepened my understanding of the Kubernetes API and its workings.

I significantly improved our code during my time, resulting in a tenfold increase in page load speed for multiple clusters. I also ensured all those controllers were covered with automated and stable tests, which are particularly challenging in Kubernetes development.

In the beginning, I worked on the Platform Automation team, which acted as a bridge between product teams and the platform. My role was to improve visibility and provide a user-friendly experience for product engineers. I achieved this by developing a Single Page Application (SPA) that connected to various platform APIs and Kafka topics. Later on, I transitioned to the Edge Team, where I focused on our API Gateway. This Gateway combines Nginx with Lua Plugins and handles tasks like API authentication and rate limiting. We automated and

managed it using Go services whenever feasible

I worked on highly scalable applications, developing and

maintaining a dozen microservices, some reaching more than 1 Million requests per minute.

At that time, I delivered a successful 2018 FIFA World Cup live stream for more than 1.3 million simultaneous users in some matches with more than 2.2 Tbps outgoing traffic.

I also drove the adoption of Kubernetes by automating the cluster deployment from scratch using Puppet, connecting the network with our BGP routers, integrating with our LDAP, CI/CD pipeline using GitLab, and so on. All that is to help us manage and deploy our apps in over 120+ bare metal servers.

Worked with Software Defined Networks (SDN) developing OpenFlow applications to HP SDN Van Controller, e.g., Firewall, Load Balancer, and Traffic Steering.

I also wrote a fault-tolerant, concurrent Ruby on Rails app, which integrated a bunch of APIs in one single panel to a centralized management. In this project we also had to write many API clients for this project. With this project, I had to write many API clients, make the system tolerant to API failures, and reduce the latency by parallelizing every network call as much as possible.

Developed Wunderlist version 2, a to-do list application

used for 3 million users.

In this project, I had to split my time between writing the front-end portion of the application using Backbone.js and their API, which was built using Ruby on Rails.