CV

This is an online version of my education, skills and work experience. You can find a shorter PDF version here.

Last updated: January 2024.

My main interests are data visualization, statistical modeling (particularly Bayesian modeling), experimental design, and predictive analytics.

Skills

Statistics, machine learning and data mining:

• Bayesian statistics
• Choice modeling
• Design of Experiments
• Data visualization
• Generalized Linear Models
• A/B testing
• Recommender systems
• Market basket analysis
• Artificial neural networks
• Random forests

Programming languages and other technologies:

• R
• Python
• C++
• SQL
• LaTeX
• Git
• Apache Impala
• Apache Hive

I also have some experience with:

Spatial Statistics, Deep Learning, Keras, Tensorflow, C, Java, MATLAB, bash, Apache Spark, Apache Hadoop, finance, linear programming, Natural Language Processing, image processing, text mining, constrained and unconstrained numerical optimization, analysis of algorithms

Work experience

October 2023 – Present day

Principal Statistician at GSK, Belgium (Full time)

Part of the Vaccines Pre-Clinical & Research Statistics team.

October 2020 – Present day

Freelance data scientist and programmer

Small freelance projects that include:

• Data visualization.
• Speed-up of slow R code by creating C++ functions on the backend.
• Analysis of Twitter data.
• Choice modeling for market research.

October 2019 – October 2023

Graduate researcher at KU Leuven, Belgium (Full time)

Research in choice modeling, experimental design and statistics in the Biostatistics group at the Mechatronics, Biostatistics and Sensors (MeBioS) division of the Biosystems Department.

I created the opdesmixr R package for computing Bayesian D- and I-optimal designs for choice models involving mixtures of ingredients and with the assumption of a multinomial logit model. The package uses C++ under the hood to maximize efficiency. Still a work in progress.

A PDF of my PhD thesis is available here.

Writing:

• Bayesian I-optimal designs for choice experiments with mixtures. Paper published in 2021 using my opdesmixr R package. The paper can be found here, and the PDF can be downloaded here. The accepted manuscript can also be found on Arxiv. To reproduce the results in the paper, go to this Github repository.
• Bayesian D- and I-optimal designs for choice experiments involving mixtures and process variables. Paper published in 2023 using my opdesmixr R package. The paper can be found here, and the PDF can be downloaded here. The accepted manuscript can be found on Arxiv.
• Why do fans go to football games? A discrete choice analysis of ticket buyers’ preferences. We modeled the preferences and willingness-to-pay for match tickets through statistical analysis of a stated-preference choice experiment performed on supporters of the Belgian national teams. The paper can be found here. The PDF of the accepted manuscript can be downloaded here.
• Finding fruit flies’ favorite color: a discrete choice experiment. Work in progress. We performed choice experiments with fruit flies and used a Bayesian hierarchical logit model to find their favorite color.

Talks, seminars and posters:

• Seminar titled I-optimal versus D-optimal designs for choice experiments with mixtures given at the DOE-IT Seminars on Design of Experiments on June 19th, 2020.
• Talk titled Bayesian I-optimal designs for choice experiments with mixtures given at the ENBIS-21 Online Conference on September 14th, 2021. A practice run of the talk can be found here, and the slides can be found here.
• Poster titled Bayesian optimal designs for choice experiments with mixtures presented at the Design and Analysis of Experiments 2021 Conference Series on October 14th, 2021. The poster can be found here.
• Talk titled Bayesian D- and I-optimal designs for choice experiments with mixtures using a multinomial logit model, given at the RSSB-21 Conference in Liège, Belgium on October 22nd, 2021. A video of it can be found here, and the slides can be found here.
• Seminar titled Bayesian D- and I-optimal designs for choice experiments with mixtures and process variables using a multinomial logit model given at the DOE-IT Seminars on Design of Experiments on May 10th, 2022 in Leuven, Belgium. A video of it can be found here.
• Talk titled Bayesian D- and I-optimal designs for choice experiments involving mixtures and process variables, given at the International Choice Modelling Conference (ICMC) in Reykjavik, Iceland on May 25th, 2022. A video of it can be found here, the slides can be found here, and the code to reproduce the results can be found here.
• Poster titled Bayesian optimal designs for choice experiments involving mixtures of ingredients and process variables presented at the Bayesian Young Statisticians Meeting (BAYSM) in Montréal, Canada on June 22nd, 2022. The poster can be found here.
• Talk titled Bayesian D- and I-optimal designs for choice experiments involving mixtures and process variables, given at the 13th Model-Oriented Data Analysis and Optimum Design (mODa13) conference in Southampton, England on July 10th, 2023.
• Talk titled Bayesian D- and I-optimal designs for choice experiments involving mixtures and process variables, delivered at the Second International Workshop of the Scientific Research Network on Choice Modelling in Leuven, Belgium on September 14th, 2023. A video of it can be found here, and the slides can be found here.

Other work:

• Thesis supervision. My PhD role includes supervising master theses from the master program of statistics at KU Leuven. As a supervisor, I provide the students with guidance on their research and code, and give feedback on their written text.
  • 2020 - 2021: The work was about finding locally D- and I-optimal designs for choice experiments involving mixtures and process variables.
  • 2022 - 2023: Design and analysis of a choice experiment involving fruit flies and mixtures of colors.
  • 2022 - 2023: Finding optimal designs for choice experiments involving mixtures with a constraint in the number of different mixtures and choice sets.
• I recreated a garden water sprinkler simulator for my supervisor to use as a project in one of his Experimental Design courses. This simulator was ported from an older implementation in MATLAB to an implementation in R which is accessible to all the students via the Shinyapps website by RStudio and can be found here.

October 2018 - October 2019

Data scientist at Banco Azteca, Mexico City (Full time)

I mostly analyzed data for the marketing and CRM teams of the bank. I was also in charge of creating and optimizing software for these analyses. Additionally, I managed and helped junior analysts in the area. I also presented results to non-technical stakeholders, such as directors and managers of the bank. Projects included:

• Visualization and analysis of large customer datasets.
• Models for the prediction of future bank transactions.
• Spatial analysis of customer and branch location data.
• Marketing campaign analysis and causal inference.
• Optimized the sampler of a Bayesian hierarchical model for BTYD analysis suited for large scale data analysis.
• Creation of marketing campaigns based on customer analysis, particularly:
  • Debit customer segmentation based on regularity of deposits and withdrawals.
  • Identification of highly regular remittance receivers from the US and Mexico.
  • Prediction model to identify remittance customers who are willing to accept a bank loan.

September 2017 - September 2018

Data scientist at Business Data Evolution, Mexico City (Full time)

Statistical modeling and predictive analysis in data science consulting company. Tasks and projects:

• Performed statistical analysis, exploratory data analysis and developed models in R.
• Created data visualization dashboards using Rstudio’s Shiny.
• Analysis of predictive models using Individual Conditional Expectation (ICE) and variable interaction quantification.
• Deployed final predictive models in Microsoft Azure VM and Azure Machine Learning.
• Developed Bayesian predictive model for small black and white documents belonging to 4 classes.
• Developed a Bayesian time series demand prediction model for a national grocery store.
• Met with clients to understand their problems and needs.
• Presented results to non-technical audiences.

May 2016 - June 2017

Data scientist at CAD Salud, Mexico City (Full time)

Created predictive and inferential statistical models, as well as data visualization dashboards for data projects aimed toward health problems.

• Performed statistical analysis, exploratory data analysis and developed models in R.
• Created data visualization dashboards using Rstudio’s Shiny.
• Created simple predictive model for potential diabetes patients in public hospitals.
• Generated spatial mathematical model for vaccines demand in public hospitals in Mexico.
• Aided in the creation of multivariate index to rank public hospitals.

October 2015 - May 2016

Data scientist at Grupo Salinas, Mexico City (Full time)

Designed and built models for predictive and inferential analysis, and created data visualization dashboards for market research consulting.

• Performed statistical analysis, exploratory data analysis and developed models in R.
• Created data visualization dashboards using Rstudio’s Shiny.
• Aided in creation of predictive model of loan origination clients in a Mexican bank.

June 2015 – August 2015

Junior Analyst at Revenue Management, Best Day, Cancún, México (Full time)

Summer project involving customer analysis and market segmentation using association rule learning (market basket analysis).

June 2014 – June 2015

Data analyst at Computer Research and Analysis Laboratory, ITAM, Mexico City (Part time)

Machine learning application in Natural Language Processing for automation in news classification for media monitoring company.

• Cleansed news articles data for modeling, mainly using R.
• Created different predictive models for automatic news classification in Python using Scikit Learn.

December 2013 – June 2014

Intern at Global Banking and Markets, HSBC, Mexico City (Part time)

Financial products pricing, financial products analysis, Excel spreadsheets creation for statistical analysis.

Education

PhD candidate at KU Leuven, Belgium: 2019 – 2024

Statistics and experimental design.

PhD thesis available here.

Computer Science MSc: 2016 - 2017

Relevant coursework:

• Machine learning
• Bayesian generalized linear models
• Bayesian multilevel modeling
• Computational statistics
• Data product architecture
• Analysis of algorithms
• Programming languages
• Complexity theory
• Advanced operating systems
• Distributed systems
• Computer architecture
• Compiler design

Note: Took around 70% of the courses in the Data Science master’s program at ITAM.

Thesis: A comparison of frequentist methods and Bayesian approximations in the implementation of Convolutional Neural Networks in an Active Learning setting. You can find the code here and the final PDF here.

Applied Mathematics BSc: 2010 - 2015

Relevant coursework:

• Statistics:
  • Probability theory
  • Mathematical statistics
  • Stochastic processes
  • Linear regression
  • Multivariate statistical methods
  • Bayesian statistics
  • Statistical learning
  • Generalized linear models
  • Spatial statistics
• Computer Science:
  • Algorithms and programming
  • Data structures
  • App development
  • Data mining
• Optimization:
  • Numerical analysis
  • Linear programming
  • Applied numerical analysis
  • Constrained and unconstrained numerical optimization
  • Parallel optimization
• Math:
  • Integral and differential calculus
  • Linear algebra
  • Real analysis
  • Dynamical systems

Thesis: Implementation of a recommender system based on matrix factorization and stochastic gradient descent. You can find the code here and the final PDF (in Spanish) here.

Languages

Spanish: Native

English: Native

French: Intermediate (B1)

Dutch: Intermediate (A2)

Hobbies and interests

Running: at least once a week for a few kilometers each time

Skiing: I enjoy this activity very much, even though I am not the best skier out there

Hiking: going to the forest or mountain with friends or colleagues

Cycling: I like long-distance road cycling with friends and team members, but I also sometimes do mountain biking

Bicycle maintenance and mechanics: I like fixing my bikes myself and keeping them in good shape

Language learning: I learned French mostly on my own, and I am currently taking Dutch lessons

Reading: mostly non-fiction titles, but also classic and modern novels

Traveling: I like traveling to different countries and cities to experience new cultures

Playing guitar: I have been playing for 18 years, and I am taking lessons to improve my technique on classical guitar

Cooking: being away from my country made me appreciate and learn the recipes of my culture

Generative art: I started learning how to combine programming and data visualization to create visually appealing images