Ph. D Candidate researching how to model relationships between entities. I'm working on how to better model and understand how entities, people, cars, Internet of Things agents, interact with each other.

Somos is a legal tech that aims to fight for the dignity of people around the globe that suffer from injustices. I'm working as a data scientist on data from class action law suites across the globe. I create insightful data analysis and deliverables, such as data products and reports, using many different technologies, including AWS Stack.》Dam Collapse Case : This case impacts over 200.000 victims in Brazil. Our team is responsible for dealing with the acquisition, data analysis, and up to delivery of reports and key insights with data visualization. My key role was to implement the data analysis of moral damages to our victims. The reports and insights I create help support the legal team in fighting for better compensation.• Stack: AWS Sagemaker, AWS Athena, AWS S3, Natural Language Processing (NLP), Pandas, Numpy, Scikit-Learn, Matplotlib, Python, SQL.》Data Product: Our team has been working on a better visualization tool for deliverables regarding our cases. Instead of delivering a huge excel table, we created a data product that shows all the analysis created and allows the user to filter, replace and create their own analysis, making it easier to understand what we want to show. My key roles are to create analysis methods that are modular and case-independent so that we can expand this tool for further cases, to support our junior data analyst when creating our data products for data quality, and coding standards, and also to brainstorm new ideas and features for the product at each sprint. • Stack: Streamlit, GIT (Bitbucket), Docker, Docker Compose, Virtual Environments, Plotly, Python.

I worked as a consultant on a money laundering detection project. My key roles were to discuss, implement, test, and evaluate how well the proposed models faired compared to the traditional model. Our most crucial deliverable was a finalized machine learning model capable of identifying clients suspicious of money laundering. The model was extremely well received and is in the production stages now. Our work earned us the best paper medal at the 2022 Conference of the Brazilian Computational Society.