Looking for a job

Nícolas Behar

@ncolasbehar

Message

I build production machine learning, causal inference, and quantitative systems that improve decisions and efficiency.

Brazil

Message

What I'm looking for

I'm looking to build and deploy reusable, well-tested machine learning and quantitative solutions for real-world decision-making, optimization, and operational impact.

I'm building and evaluating production AI and quantitative systems across healthcare, finance, trading, credit risk, and cloud optimization.

At Hapvida NotreDame Intermédica, I lead Difference-in-Differences and Synthetic Control evaluations for AI models, measuring impacts on medical staff productivity, error rates, and adoption. I also standardized experiments for clinical computer vision and built systems to flag medical misclassifications associated with increased patient risk.

At Deméter Agro-solutions, I cut model-training and backtest processing time by up to 70%, moving workflows from days to hours while refactoring monolithic research code into a maintainable, tested multi-file structure with zero production incidents. Previously at Nubank / NuInvest, I helped reduce AWS costs for critical services by more than 50%, built causal marketing models, and developed Spark data pipelines and time-series forecasts.

I've also deployed monitoring and alerting solutions for AML, risk, credit, and sales at Banco BS2, and developed credit-scoring models and data pipelines at Guiabolso. I bring a Financial Engineering master's foundation and hands-on experience with Python, Spark, Scala, SQL, machine learning, econometrics, and systematic trading research.

Experience

Work history, roles, and key accomplishments

Current

Data Science Tech Lead

Current

Hapvida NotreDame Intermédica

Aug 2026 - Present (1 month)

Led the design of Difference-in-Differences and Synthetic Control evaluations for AI models scheduled for deployment, establishing frameworks to measure effects on medical staff productivity, error rates, and adoption.

Difference In Differences Synthetic Control Causal Inference AI model evaluation Experimental Design

Data Scientist

Hapvida NotreDame Intermédica

Mar 2026 - Jul 2026 (4 months)

Conducted and standardized targeted experiments on clinical computer vision models to identify failure modes and gaps in chest X-ray disease detection. Developed a system to flag medical misclassifications and surface combinations of exam alterations associated with increased patient risk.

Computer Vision Clinical AI Experimental Design Medical Imaging Risk Assessment

Quantitative Analyst

Deméter Agro-solutions

Aug 2025 - Jan 2026 (5 months)

Led a critical infrastructure initiative to improve model training and backtest code, reducing processing time by up to 70%. Mentored other analysts and refactored feature storage architecture to improve data retrieval and processing.

Python Machine Learning Backtesting ETL Git YAML Performance Profiling Mentorship

Data Scientist, Decision Science

Banco BS2

Apr 2024 - Apr 2025 (1 year)

Developed and deployed data-driven alerts for FX, payments, and betting-related activity in collaboration with AML teams and risk leadership. Led technical development of monitoring and alerting solutions and automated ETL processes.

Data Science AML Risk ETL Data Warehouse Data Lineage Monitoring Alerting

Data Scientist

Nubank / NuInvest

Nov 2020 - Mar 2024 (3 years 4 months)

Supported AWS cost optimization by building models that recommended optimal machine counts, reducing costs by over 50%. Developed causal inference models for marketing campaigns and contributed to internal tooling and open-source FK Learn.

Python Spark SQL AWS Causal Inference Machine Learning NLP Time Series ETL Open Source

Business Intelligence Analyst, Data Scientist

Guiabolso

Sep 2017 - Oct 2020 (3 years 1 month)

Researched, developed, and monitored machine learning models for credit scoring. Built data pipelines and ETL processes, and automated data-collection processes saving up to 4 hours per day.

Machine Learning Credit Scoring ETL Data Pipelines Data Analysis Automation