Teaching Assistant
2020 - PRESENTUniversidad Católica Argentina- Composed practical lessons for the Data science class.
- Corrected exams and practical homework for the Data science subject.
- Tutored and answered questions from students in the Data science topics.
- Explained machine learnning techniques, regularizations methods, feature extraction and selection, data visualizaton methods, and more tasks related to data science.
Technologies: Data SciencePh.D. Student Researcher
2017 - PRESENTUniversidad de Buenos Aires- Created new techniques to analyze discussions on social networks with R and Python.
- Predicted movie reviews using the IMDB database (R).
- Predicted implication clauses with NLP through Python models.
- Developed new techniques to predict controversy with NLP techniques on social networks with R and Python.
- Created new techniques to graph clusters with NLP techniques on social networks with Python and R.
Technologies: Python, RFreelance Data Scientist Advisor
2016 - PRESENTMassomedia S.A.- Predicted presidential votes via telephone surveys using Python.
- Analyzed several discussions on Twitter and Facebook using R.
- Developed a web app using R, PHP, and MySQL to monitor social networks and media.
- Built a product with Python that could analyze telephone surveys about product sales.
- Presented the results, conclusions, and explanations for each task to the client.
Technologies: R, PythonStatiscal Developer
2020 - 2020Decentral Park Advisors LLC (via Toptal)- Predicted Bitcoin 1D, 3D, and 7D returns with multinomial regressions and GAM models.
- Predicted Bitcoin 1D, 3D, and 7D positive or negative values through classification models as RandomForest, XGBoost, and Bagging.
- Reported the results through Jupyter Notebooks and R Shiny dynamic graphs.
Technologies: Python, Sklearn, Pandas, Matplotlib, Jupyter, R, ShinyData Analyst
2020 - 2020LL Media, LLC (via Toptal)- Standarized multiple information sources about leads using Python and Pandas.
- Scored the performance of each source over different kind of campaigns using Python, Pandas, and Matplotlib.
- Predicted good leads by demographic data using machine learning classifiers (sklearn).
- Predicted bad leads by demographic data using machine learning classifiers (sklearn).
- Analyzed lead data to find simple correlations between good and bad lead performance.
Technologies: Python, Sklearn, Pandas, MatplotlibTeaching Assistant
2017 - 2018Universidad de Buenos Aires- Composed practical lessons for the Computer Structures 1 class.
- Corrected exams and practical homework for the Computer Structures 1 class.
- Tutored and answered questions from students in the Computer Structures 1 class.
Technologies: Computer StructureTeaching Assistant
2016 - 2017Universidad de Buenos Aires- Composed and gave practical lessons for the Network Theory class.
- Corrected exams and homework for the Network Theory class.
- Tutored and answered questions from students in the Network Theory class.
Technologies: Network TheorySenior Front-end Developer
2014 - 2015Telam- Developed and maintained a system for journalists which allowed them to write different types of notes and publish them into the news site.
- Built and maintained a system to administrate the advertisement money.
- Developed a REST API to connect with other media news sites.
Technologies: PHP, MySQL, JavaScriptSenior Front-end Developer
2010 - 2014Intraway- Built and maintained a system for callcenter assistance.
- Developed and maintained features to communicate with the STB systems and reset them.
- Constructed and supported a dynamic decision tree to give the best answers to clients based on their specific problems and configurations.
- Created and maintained an internal ticket system to organize tasks and assign them to different teams.
Technologies: PHP, SQL, JavaScript, jQueryPrincipal Developer
2008 - 2010Imprek- Developed the company's administrative system.
- Built a system to administrate technique services.
- Maintained the stock system.
- Implemented a system to print digital photos from a Kodak machine.
- Developed the company's financial system.
Technologies: PHP, MySQL, JavaScript
Machine Learning Developer in Ciudad de Buenos Aires, Buenos Aires, Argentina
Member since October 2, 2019
Currently, Juan is a Ph.D. candidate a the University of Buenos Aires, researching the subjects of AI, NLP, and social networks. He has over a decade of professional development experience under his belt. For the last few years, he’s been immersing himself in various types of data science projects and loving every minute of it. Juan relishes taking on data problems, building prediction models, and learning state-of-the-art techniques.
Juan is now available for hire
Portfolio
- Universidad Católica ArgentinaData Science
- Universidad de Buenos AiresPython, R
- Massomedia S.A.R, Python
Experience
- Data Science 5 years
- Machine Learning 4 years
- RStudio 3 years
- Python 3 3 years
- Jupyter Notebook 3 years
- Pandas 2 years
- Matplotlib 2 years
- Sklearn 2 years
Availability
Part-time
Preferred Environment
Jupyter Notebook, RStudio
The most amazing...
...thing I've coded is a an entire web app to monitor social networks. It has a back end in R for statistics and a PHP front end.
Employment
Experience
- Application to Monitor Social Networks (Other amazing things)http://sonar.monteagudo.com.ar/
I developed, on my own, an entire application to monitor social networks like Instagram, Facebook, and Twitter. It has statistics about any public account needed and also sends messages/alarms to the client through a telegram if something important is happening.
It has a back end in R to download information, process it, and calculate statistics and a PHP front end to show the data. I also used C to manage sessions, create, delete and modify searches, and more. - Stars Realigned: Improving the IMDb Rating System (Other amazing things)https://www.toptal.com/data-science/improving-imdb-rating-system
IMDb ratings have genre bias: Dramas tend to score higher, for example. Is there a way to remove such biases and discover what makes a movie unique?
In this article, I show you how to refine IMDb scores and create a better ranking system through data science and machine learning techniques. - Stars Realigned: Improving the IMDb Rating System (Publication)IMDb ratings have genre bias: For example, dramas tend to score higher. Removing common feature bias and keeping unique characteristics, it's possible to create a new, refined score based on IMDb information.
Skills
Languages
Python 3, R, SQL, JavaScript, CSSFrameworks
RStudio Shiny, CodeIgniterLibraries/APIs
Sklearn, Matplotlib, Pandas, Keras, Ggplot2, NumPy, jQuery, CaretParadigms
Data Science, Test-driven Development (TDD)Platforms
RStudio, Jupyter Notebook, Linux, WordPress, OracleStorage
MySQLOther
OOP Designs, Machine Learning, Data Analytics, Data Analysis, Big Data, StatisticsTools
Dplyr, Seaborn
Education
- Ph.D. degree (in progress) in Computer Science2017 - 2022Universidad de Buenos Aires - Buenos Aires, Argentina
- Master's degree in Computer Science2010 - 2016Universidad de Buenos Aires - Buenos Aires, Argentina
Certifications
- Deep LearningDECEMBER 2018 - PRESENTCoursera
- Machine LearningOCTOBER 2018 - PRESENTCoursera
- Laboratory of Machine LearningSEPTEMBER 2018 - PRESENTITBA | Instituto Tecnológico de Buenos Aires
