I'm a data generalist with a passion for practical solutions to complex problems. I excel in communicating technical insights across different levels of an organization and have a proven track record designing and building data solutions from scratch. Never satisfied with doing just one thing. Dual citizen of the US and Brazil.

Skills

Data Engineering

  • Python
  • Bash
  • SQL
  • Postgres
  • BigQuery
  • Redshift
  • Web Scraping

Data Science

  • Data Analysis
  • Data Visualization
  • Machine Learning
  • Generative AI
  • Stable Diffusion
  • Experimentation

Cloud Tech

  • AWS (S3, EC2, RDS, Lambda, ECS)
  • Google Cloud (BigQuery, GCS)

Work Experience (5)

Oct 2022 - Current
Data Engineer
IPinfo
IPinfo is a leading provider of IP address and internet data. As a hybrid data engineer and scientist, I work to build the pipelines that power our data products and to analyze and visualize the complex and massive datasets we collect to develop new products and ensure we deliver the best, most reliable internet data that exists.
  • Built and maintained pipelines processing over 30TB of data per day using BigQuery, Bash, and Python
  • Developed new products and ensured data quality by analyzing and visualizing complex and massive datasets
  • Improved geolocation data for hundreds of millions of IPs by analyzing and processing device data
Feb 2023 - Nov 2023
Senior Data Science Consultant
Limehome
LimeHome is a hospitality startup that offers fully equipped apartments for short-term rental in Europe. I was contracted for my expertise in building pricing models and data pipelines to support the company's rapid growth.
  • Optimized revenue and occupancy management by building a new pricing system from scratch, using a bespoke algorithm
  • Ensured stakeholder satisfaction by working directly with revenue management to ensure the design could take in their input
Apr 2021 - Sep 2021
Teaching Assistant
DS4A Empowerment
I was a Teaching Assistant for the second cohort of DS4A Empowerment, a 12-week intensive program taught by Harvard's Prof. Natesh Pillai dedicated to training around 1000 people from underrepresented backgrounds in data literacy, analytics and programming.
  • Assisted students with their lectures and projects, and conducted workshops in topics such as Python, SQL, statistics and data analysis
  • Personally responsible for guiding a cohort of 50 fellows from diverse backgrounds and nationalities
Aug 2020 - Oct 2022
Data Scientist -> Data Science Manager
Casai
Casai was a Mexican hospitality startup focused on short-term rentals in LATAM. As Data Scientist and eventually Data Science Manager, I was responsible for building the company's data infrastructure and developing pricing models to optimize revenue and occupancy. I also worked with operations, marketing and revenue management to ensure the data solutions I built were aligned with their needs.
  • Managed a team of six data analysts, scientists and engineers to ensure the company's data needs were met
  • Developed automated pricing systems that outperformed the competition and allowed for rapid expansion into new markets
  • Developed novel inventory management systems to reduce costs with amenities and inventory loss by over 50%
  • Designed, executed and tracked controlled experiments to foster a data-driven decision-making culture in the company
Nov 2018 - Jul 2020
Data Analyst -> Data Lead
Distrito
Distrito is one of Brazil's most important innovation hubs and the premier source of data on Startups in Brazil. As Data Lead at Distrito Dataminer, I built and scaled the data collection efforts that allowed us to provide the broadest and most accurate data on the startup ecosystem in LATAM.
  • Built the most comprehensive database of startups in Brazil, with data on over 13,000 companies, by automating data collection and processing
  • Led the production of market intelligence reports and startup maps that took Distrito from 0 to 160 citations in the press in 2019
  • Created methods that allowed Distrito to predict 4 out of the 5 new Brazilian Unicorns in 2019
  • Presented Distrito Dataminer's findings at several events including the IV National Commerce Forum, to over 900 people.
  • Did essential work in infrastructure and operations that allowed Dataminer to quickly evolve from an area with no revenue to one that earned over R$2 million in 2020

Projects (4)

Letterboxd.tools
 https://letterboxd.tools/
Toolboxd is a suite of tools for users of Letterboxd, a social network for film lovers. It includes a recommendation engine, random movie picker, and a film taste compatibility test.
  • Built a recommendation engine trained on hundreds of millions of reviews that uses collaborative filtering to provide excellent personalized movie recommendations
  • Got immensely positive feedback and over 20,000 users in total so far, with a few hundred interactions daily
  • Led to a freelance project with Letterboxd themselves to provide insights from their data to distributors and studios
  • Streamlit / Python / Postgres
Multiversions
 https://multiversions.com.br/
What started as a Halloween project to use Stable Diffusion to create monsterized versions of my friends became a side project that can generate AI images of any subject in a variety of styles.
  • Fully custom and automated Generative AI pipeline to train, deploy and run inference on Dreambooth fine-tunes of Stable Diffusion
  • Able to learn the features of people and pets and apply dozens of different styles to them with high quality
  • Developed solution to curate and filter results so that customers get the best possible images
  • Python / AWS Lambda / Runpod / Postgres
NBA BetBot
Worked with a professional sports bettor in Australia to develop a bot that would live-trade NBA game odds on Betfair.
  • LightGBM / XGBoost compound ML model to predict game outcome probabilities live with a slight edge over the market
  • Infrastructure to automatically collect data, make bets and manage the bankroll
  • Beat the market, but couldn't beat the commission
VTEX Store-Agnostic Recommendation System
VTEX is a major e-commerce platform. As a capstone project for DS4A Latam, I developed a collaborative filtering recommendation system that could be used by any store on the VTEX platform, regardless of their size or type.
  • Runner-up grand prize/People's choice winner in the DS4A Latam Capstone competition
  • Presented the project to the VTEX executive board

Education (3)

2013 - 2013
Bachelor's
 Astronomy (incomplete)
University of São Paulo
2014 - 2018
Bachelor's
 Communications
University of São Paulo
2020 - 2020
Specialization
 Data Science
DS4A Latam by Correlation One