Scroll To View More
João Filgueiras, Machine Learning Developer in Porto, Portugal
João Filgueiras

Machine Learning Developer in Porto, Portugal

Member since February 22, 2014
João is an accomplished researcher and prototype developer with a special talent for breaking down large problems into solvable pieces. His background in R&D gives him the edge in implementing efficient, effective, and sophisticated yet clean solutions.
João is now available for hire

Portfolio

Experience

  • JavaScript, 11 years
  • Python, 6 years
  • Highcharts, 5 years
  • Node.js, 4 years
  • Machine Learning, 3 years
  • Data Visualization, 3 years
  • Scikit-learn, 2 years
  • Vue.js, 2 years
Porto, Portugal

Availability

Part-time

Preferred Environment

Linux, Atom, Git, Jupyter

The most amazing...

...thing I've coded is a machine learning classifier to classify opinions in a text as positive/negative/neutral, and to infer personal attributes of its author.

Employment

  • Invited Assistant Professor

    2019 - PRESENT
    University of Porto
    • Co-lectured a course on algorithm analysis and design for second year undergrad students.
    Technologies: Algorithms
  • AI Researcher

    2019 - PRESENT
    Artificial Intelligence and Computer Science Laboratory – University of Porto
    • Worked on a public interest project with the aim to modernize and optimize government agencies using AI and data science.
    • Used NLP techniques to extract useful information from unstructured text.
    • Explored a large and complex public service dataset.
    • Worked on classification problems using machine learning techniques.
    Technologies: Python, Pandas, Jupyter, NLP, Machine Learning
  • Senior Software Engineer

    2018 - 2019
    Undisclosed Fintech Company
    • Implemented automated trading strategies for cryptocurrency markets.
    • Designed the system architecture to ensure stability and fault tolerance.
    • Implemented custom data visualization pieces.
    Technologies: Node.js, React, HighCharts
  • Technical Screener for AI and Data Science

    2018 - 2019
    Toptal, LLC
    • Interviewed potential AI and data science experts looking to join the Toptal network.
    • Mentored other AI and data science screeners.
    • Contributed to the creation of the screening process and continuously refined it.
    Technologies: Internal Tools
  • Machine Learning Expert

    2018 - 2018
    Rainbows (via Toptal)
    • Performed a thorough exploratory analysis of a tabular dataset.
    • Researched resource allocation schemes based on a number of metrics and constraints.
    Technologies: Python, Jupyter, Pandas
  • Director of Engineering

    2017 - 2018
    Toptal, LLC
    • Performed a key operational role within the company, as part of a team, by understanding client requirements and matching them with the best talent in the network.
    • Helped shape and improve the operational processes of the company.
    • Co-developed and tested new screening processes and development specializations to ensure the highest standards of talent quality in new areas.
    • Managed several accounts for SMB and enterprise clients all over the world.
    • Supported and advised talent working with clients.
    Technologies: Internal Tools
  • Lead Web Developer

    2016 - 2017
    ThinkProcess
    • Implemented a web-based business process management system that allows management, visualization, and reporting of business processes and supporting documents.
    • Developed the front-end using ES6 and Vue.js 2.x with a focus on component re-usability.
    • Used Node.js and MongoDB for the back-end.
    • Built complex data models with MongoDB, including tree structures.
    • Integrated and customized a BPMN diagram editor.
    Technologies: Vue.js, Node.js, MongoDB
  • Technical Screener

    2015 - 2017
    Toptal, LLC
    • Interviewed hundreds of applicants looking to join the Toptal network and ensured their technical skills, work ethic and communication skills met Toptal's standards.
    • Mentored other technical screeners.
    • Helped improve the screening process.
    Technologies: Internal Tools
  • Full-stack Developer/Data Scientist

    2014 - 2017
    Agri-Esprit, SA
    • Developed multiple front-end pages using JavaScript with the Maria MVC framework.
    • Implemented a framework for building complex agronomic models to evaluate crop status and recommend suitable actions.
    • Evaluated a big data storage and data mining tool to work with massive collections of data in the domain of large-scale agriculture.
    • Implemented a framework for web front-end testing using CasperJS, specific to the company's software product.
    • Worked on data visualizations and a process management editor using the BPMN notation.
    Technologies: JavaScript, Node.js, CasperJS, jQuery, Python, Pandas
  • Data Scientist/Data Visualization Expert

    2014 - 2014
    nuOctave
    • Implemented data processing and visualization tools for advanced pattern finding in financial market data using HighCharts and JavaScript on a web app.
    • Created several data visualizations for financial market data on a web app using HighCharts and JavaScript.
    • Implemented a fully featured graphical widget to allow a user to build a logical query by wiring a given list of base conditions and logical nodes such as AND/OR, using just basic SVG shapes and D3.js.
    • Implemented interactions with several different APIs to build interactive front-end sections of the web app.
    • Implemented a Python application to explore and analyze pattern data.
    Technologies: JavaScript, D3.js, HighCharts, DC.js, Python
  • Researcher/PhD Candidate

    2012 - 2014
    INESC-ID
    • Developed a state-of-the-art machine learning approach to sentiment analysis using Python and sklearn.
    • Developed of a web page for data visualization and dissemination using PHP, jQuery, and HighCharts.
    • Created a state-of-the-art machine learning-based approach to infer demographic attributes of the author of a given text using Python and sklearn.
    • Co-developed a web page to showcase different text-mining tools using Python, Bootstrap, jQuery, and D3.js.
    • Co-developed rich data visualizations for interconnected political mandates using D3.js.
    Technologies: Python, Machine Learning, PHP, jQuery, D3.js
  • Undergraduate Researcher

    2009 - 2012
    IT Porto (via Carnegie Mellon University - Portugal Programme)
    • Worked on a project designed to collect and analyze vehicular traffic data based on Bluetooth devices.
    • Created data analysis tools for Bluetooth data using Python.
    • Planned a city-wide deployment of Bluetooth traffic scanners.
    • Developed the collection application in Python/C as well as the back-end to receive and store remote Bluetooth logs using PHP.
    • Developed web-based tools for data visualization of geographic data using Google Maps, PHP, and JavaScript.
    Technologies: C, Python, PHP, Bluetooth, JavaScript, Google Maps

Experience

  • Accrue WebApp (Development)

    Developed data visualizations, filters, and front-end data processing for a sophisticated and innovative web app that allows the user to look for patterns and trends in financial market data. This project required heavy knowledge of HighCharts/Highstock, DC.js, Crossfilter, D3.js, jQuery, and Python for auxiliary tools.

  • POPSTAR Website (Development)
    http://popstar.pt/?lang=EN

    Developed and co-designed the POPSTAR website showing the trends of the Portuguese public opinion on politics. The website was developed using PHP, jQuery, and heavily customized HighCharts.

  • Author Profiling Classifier (Development)

    Developed a machine learning classifier that, using solely the features of a given text, predicts the gender, age group and political affiliation of its author. This was also developed within the text classification framework using Python and scikit-learn.

  • Sentiment Analysis Classifier (Development)
    http://popstar.pt/extras/opinionizer_poster.pdf

    Developed a sentiment analysis classifier (i.e. given a text determine whether it expresses a positive, neutral, or negative opinion) that ranked 3rd in a scientific competition with over 40 participants. It was implemented using Python and scikit-learn.

  • Text Classification Framework (Development)

    Developed a text classification platform in Python containing several thousand lines of code implementing a variety of state-of-the-art algorithms to parse, pre-process, extract features, and train machine learning classifiers from textual documents.

Skills

  • Languages

    Python, JavaScript, ECMAScript (ES6), Clojure, TypeScript
  • Frameworks

    Express.js, ClojureScript, Bootstrap, Angular
  • Libraries/APIs

    Vue.js, Pandas, Node.js, jQuery, Highcharts, DHTMLX, Scikit-learn, D3.js, jQuery UI, React, Matplotlib, Raphaël, PhantomJS, LeafletJS
  • Tools

    Jupyter, Mongoose, Git, Reagent, Jira, Atom, Sequelize, IPython, Seaborn, CasperJS, Apache Solr
  • Paradigms

    Data Science, REST, BPMN
  • Other

    Natural Language Processing (NLP), Machine Learning, Data Visualization, Bluetooth, Information Retrieval
  • Platforms

    Linux
  • Storage

    MongoDB, PostgreSQL, MySQL

Education

  • Master's degree in Computer Science
    2005 - 2012
    University of Porto (Faculty of Sciences) - Oporto, Portugal
I really like this profile
Share it with others