Description

I help businesses turn messy data into clear insights and unexpected opportunities. With a mix of analytical rigor and creative problem-solving, I uncover patterns, tell stories your numbers are hiding, and craft solutions that actually work. I typically work on predictive modeling, time series analysis, and advanced analytics projects—but I’m most interested in challenges that let me combine data science with inventive thinking. Remote, B2B, and ready to tackle problems that need more than just code.

Languages

Portuguese
Native or bilingual
English
Native or bilingual
Spanish
Conversational
French
Conversational

Workplace preferences

Remote only

Primarily works remotely

Trustly
Senior Data Analyst
May 2022 - September 2024 (2 years and 4 months)
Vitoria, Vitória - State of Espírito Santo, Brazil
Built a fully parameterized fraud alerting framework that replaced an inefficient SQL/dashboard process. Reduced investigation time per alert from 90+ minutes to under 5, integrated with Salesforce, and enabled tracking of historical alerts.
Advocated for and executed the first use of Python in the analytics org, despite cultural and tooling resistance. Later ported the alert framework to PySpark to enable automation via the data platform.
Served as analytics focal point for both Engineering and Risk, leading the full migration of dashboards from Tableau to QuickSight—the only team to complete it on time.
Led the analytics team in the company-wide documentation revamp, creating the standard dashboard and KPI templates, and training the company (including C-levels) in live sessions.
Mentored analysts and promoted Python literacy within the data team through hands-on coaching and pair programming.
Conducted a large-scale psychographics enrichment initiative with the product team, merging external data to explore persona patterns and social strata across users.
Universidade Federal de Santa Catarina,
Undergrad/Grad researcher
March 2009 - March 2016 (7 years)
Florianopolis, SC, Brazil
Led a 5.5-year research project on the eclipsing binary system V2051 Oph, resulting in 3 first-author papers and international presentations.
Built a full data pipeline (C, Bash, IRAF) to transform raw telescope images into calibrated, structured time series.
Developed custom statistical models to separate and analyze overlapping signals from stellar components using geometric eclipse modeling.
Applied Monte Carlo and bootstrap simulations to estimate parameter uncertainty (e.g., temperature, inclination, disk structure).
Produced the first photometric detection of spiral features in a quiescent accretion disk, with distance estimates later confirmed by Gaia within 1σ of model predictions.
Mentored junior researchers in scientific coding, data visualization (Gnuplot), and reproducible analysis in LaTeX.
Trustly
Senior Data Scientist
May 2025 - Today (1 year and 1 month)
Vitoria, Vitória - State of Espírito Santo, Brazil
Created the company's first persona clusterization using enriched bank activity data to segment customer behavior. This was presented to a former client during re-engagement and directly contributed to recovering Trustly's largest client.
Designed a behavioral risk scoring and segmentation system to replace tenure-based risk tiers.
The new model identified ultra-VIP customers (0.1% of users) contributing 24% of transaction value with negligible risk. Simulations showed the potential to increase revenue by 10% and reduce return loss by 22% through optimized rule targeting.
Continued mentorship and internal advocacy of modeling-driven strategies across the risk, product, and analytics teams.