Hi, I’m Wlademir Ribeiro Prates.
I’m a senior data scientist and R/Shiny engineer. Today I lead an R/Shiny development and QA team at Halo Media, building analytical applications for Marsh. Before that I spent close to a decade turning data into products and decisions, first as a consultant in Brazil and then as an engineer for companies in the United States and Europe. I hold a PhD and a Master’s in Finance from UFSC, where I researched behavioral finance and capital markets.
This site is where I collect what I learn: applied analyses with code, notes on taking data applications to production, and reflections on building a career in data.
Education
PhD in Finance
Master’s in Finance
Bachelor’s in Business Administration
Experience
2025 – present · New York, remote
Technical Team Lead · Halo Media
Joined as Senior R/Shiny Engineer and now lead the development and QA team that builds analytical tools and dashboards for Marsh. Architecture decisions, code review, automated testing and CI/CD, and mentoring developers.
2024 – 2025 · Los Angeles, remote
Senior Data Scientist, R/Shiny Developer · Firebird
Analytical web applications for the music industry, helping bands plan and optimize tours. Shiny on top of BigQuery, with rhino, R6 and telemetry.
2021 – 2024 · Warsaw, remote
Senior R/Shiny Developer, then Tech Lead · Appsilon
Shiny applications for Genmab, comparing patient treatments with statistical tests and rich visualizations. Also apps for Warner Bros., from budget planning to an internal GPT tool for film concepts.
2020 – 2021 · São Paulo, remote
Data Science Specialist · Tembici
Led analyses with statistical modeling and machine learning, and built and maintained the customer churn model in production. The same kind of work behind this article.
2018 – 2020 · São Paulo
Senior Data Scientist, then Data Science Manager · EY
People Analytics consulting for companies such as Vivo (Telefónica), BASF and Braskem: predicting employee turnover, understanding what drives productivity, and in-company training on hypothesis testing, visualization and machine learning.
2017 – 2018 · Florianópolis
Data Scientist · Aquarela Advanced Analytics
Predictive and prescriptive projects with dashboards in R/Shiny for Votorantim, Embraer and DHL, from predictive maintenance for aircraft parts to the best delivery time in logistics.
Earlier
Knowledge management in personal and corporate finance at Par Mais, and a year as professor at UFSC’s Department of Production Engineering.
Teaching
Teaching has run alongside the industry work the whole time. Since 2024 I teach online data science classes and mentor people starting in the field at Data Mundo. From 2019 to 2021 I taught the machine learning and NLP modules of an MBA in artificial intelligence and big data at Sustentare Escola de Negócios, and in 2015 I was a professor at UFSC.
What I work with
- R and Shiny, from exploratory analysis to applications used in production, with testing, CI/CD and telemetry around them.
- Machine learning applied to business problems, with attention to whether the model actually changes a decision.
- SQL and cloud data warehouses such as BigQuery and Redshift, plus Python and LLM tooling when the problem asks for it.
- Reproducible reporting with Quarto and R Markdown.
- Finance and behavioral economics as the analytical background behind much of my work.
Book
I wrote O Manual do Cientista de Dados, a free and open handbook in Portuguese about data science in practice: statistics, modeling, visualization and the reasoning behind them. It is an ongoing project, with source and contributions on GitHub.
Research
Before moving to industry, I did academic research in behavioral finance and capital markets at UFSC, mostly on the disposition effect and how investors decide under risk. Selected papers:
- Dorow, A., da Costa, N., Takase, E., Prates, W. R. & Da Silva, S. (2018). On the neural substrates of the disposition effect and return performance. Journal of Behavioral and Experimental Finance.
- Prates, W. R., da Costa, N. & Dorow, A. (2017). Risk aversion, the disposition effect, and group decision making: an experimental analysis. Managerial and Decision Economics.
- Prates, W. R., da Costa, N., Armada, M. J. R. & Da Silva, S. (2019). Propensity to sell stocks in an artificial stock market. PLoS ONE.
- Prates, W. R., da Costa, N. & Santos, A. P. (2019). Efeito disposição: propensão à venda de investidores individuais e institucionais. Revista Brasileira de Economia.
The full list is on Google Scholar.
Get in touch
The best place to reach me is LinkedIn. For shorter, more informal content, I’m on Instagram and YouTube.