Mariana Meireles

Currently, I am working on developing an internship program aimed at empowering junior developers from the global south to contribute to projects within the NumFOCUS ecosystem. In addition to this, I am actively involved in organizing the NumFOCUS DISC Unconference 2023.

Furthermore, I am contributing to shaping a new edition of the NumFOCUS DISCOVER Cookbook.

In the PyScript project, I have been involved in both developing code and building community. I'm responsible for kicking off the PyScript Collective effort alongside Paul Everitt. I have also joined colleagues and the broader Python community in order to coordinate efforts to create a centralized set of protocols for working with Python on the web.

Technologies: Python, WASM, JavaScript, TypeScript

I have worked with the Ethos of Open Science team, where I assisted in the development of a program focused on leveraging open science practices. In order to create this program, I conducted research on the current best practices and recommendations for working in the open. This program is being converted into an EDX course.

I was invited by the LSSTC-DSFP program to teach Astronomy graduate students on topics such as Object-Oriented Programming (OOP), performance optimization for scientific software, open science, and reproducible research practices. While the classes took place at Harvard University, I was an invited guest lecturer and not a part of Harvard's faculty roster.

Integrated the Cytoscape to the Jupyter Ecosystem as a widget[1]. Created C++ kernels to run SQLite[2] and later on general SQL code[3] with data visualization built-in in the Jupyter Notebook. Developed a package manager for R that's 25x faster than CRAN[4] and a deprecation finder for JupyterLab that runs on static code based on Memestra[5]. I also wrote an article on different jupyter notebooks visualizations[6] and contributed with a few new features. Extensively contributed to Mamba, an OS agnostic package manager written in C++.

[1] Interactive Graph Visualization in Jupyter with ipycytoscape
[2] A Jupyter kernel for SQLite
[3] An SQL solution for Jupyter
[4] Rhumba: a faster R distribution
[5] Memestra!
[6] Looking at notebooks from a new perspective

Technologies: Python, C++, Typescript, Javascript, CMake, CI, git, Multiplatform

Responsible for maintaining and building new features for PySide (the Python bindings for Qt). Developed the "Default" menu and templates for PySide on QtCreator. Addressed bugs and worked on overall housekeeping activities in the PySide repository and community. I have also written documentation and created a Qt to PySide code converter, a tool that automatically translates documentation from Qt to PySide2.

I published an article in the official Qt page about a demo I've developed to show the interaction with PySide, QML and Python machine learning libraries in the QtCon Berlin - 2019.

You can check part of my contributions here.

Technologies: Python, C++, QML, NLP, git, Jira

Cut the average loading time of the Webcompat website by 35% by using server-side rendering, HTTP/2 and better compression algorithms.

Using tests designed on Selenium I was able to track down some performance bottlenecks and decrease the page load time by 1s in low-speed mobile connections.

Removed embedded JS libraries that weren't necessary while making some small changes to the code base to keep the same behavior.

Worked on some small patches for Firefox that are available on my bugzilla profile.

Technologies: Backbone, Jinja, NodeJS, Backbone, Flask, nginx, Jquery, git, Phabricator, Python, Javascript, Selenium

At the company, I successfully integrated our chatbot system with various WhatsApp APIs, both unofficial and official. I enhanced the chatbot by adding new, user-friendly features and developed a comprehensive metrics dashboard, including backend implementation. Additionally, I improved the user experience by implementing pagination, significantly reducing the loading time of multiple pages.

Technologies: Python, Javascript, NodeJS, Flask, Bootstrap 3, AngularJS, Express, Docker, MySQL, Agile methodology, MVC

As an electrical engineering student with a focus on computer science, I worked with Dr. Flavio Figueiredo to conduct research using data extracted from Github's Google BigQuery profile. Our goal was to explore potential correlations between the patterns of interaction among users in open source communities and the languages they used and were exposed to.

Technologies: Python(numpy, matplotlib, pandas), Sentiment analysis tools(VADER), data scraping

Responsible for maintaining the department's website, servers and network.

Technologies: Ubuntu server, Wordpress, Network

As an Embedded Systems Engineer at Autobotz UFMG (the university's robotics competion team), I contributed to the development of several libraries for controlling Arduinos, Raspberry Pis, and BeagleBoards in our competition robots.

In addition to hardware control, I designed and developed the team's website using Hugo, creating a polished and engaging online presence.

Technologies: C, Python, Go, Hugo