Personal website built with Gatsby
Advent of Code 2020 in Elixir (by a newbie).
Trefle API Wrapper for Elixir, built for ease of use, with some cool helper methods to easily handle pagination.
Several encryptors (cryptographic secure or not) available in a REST API fashion, using Flask
Website with all Github emojis, dynamically updated
FINNger is a CNN intended to detect how many raised fingers you have through your webcam.
Few stuff that I've learned through the years, heavily focusing on backend technologies, but with a lot of curiosities on new trends (like React and Dart).
Currently working at Leadsimple, a CRM and Operations management software purposefuly built for property managers, helping to sustain a brand new React UI backed by a Rails + GraphQL API backend, with full test coverage (RSpec) and a fully functioning CI/CD workflow connected with Heroku. Also, working on a new software built with Elixir + Phoenix to handle the synchronization between the company data and external property management softwares.
Junior Backend Enginner at TAG Livros, developing solutions for the company internal workflow. Working with frameworks such as Django, NodeJS and React, interacting consistently with AWS Tools for Serverless Computing such as S3, API Gateway, Lambda, SQS, RDS and DynamoDB. Responsible for building new internal projects such as a common interface for the company bots (Slack and Discord), as well as a CLI to enable faster production deployment of the company digital products.
Researched about Machine Learning during my undergraduation , more specifically about Graph Neural Networks and their applications. In the laboratory, worked with Python (Tensorflow and PyTorch) to achieve the high paralelism needed to train models in a satisfatory time (it's still weeks), using NVIDIA graphic processing units. Also focused on data analysis and community identification.
Developed platform for the Brazilian Health Ministry for a cooperation between them and UFRGS. The plataform was used in 2 fronts: Help solving companies problems when trying to computerize the brazilian UBS; and being able to hold a plataform responsible for solving complaints between companies, cities and UBS's. Using Rails 5 and Docker on Linux, using the MVC software engineering model and some modern techniques to boost produtivity, the team achieved to build a full web app in around 6 months.
Helped new Computer Science undergratues to complete the subject challenges each week, using C language. With this, acquired a big knowledge in algorithms, such as sorting, hash, search, data structures, git, console manipulation, etc...