Global Fraud Strategy Data Manager
2022 - PRESENTKlarna- Identified a flaw with the company's authentication system and proposed a new algorithm that doubled the acceptance and conversion rates.
- Detected and fixed key data integrity issues impacting models, analysis, and indicators.
- Mentored and trained data scientists and analysts.
Technologies: Python 3, Linux, Data Science, Data Analysis, Machine Learning, Python, Time Series Analysis, Amazon Web Services (AWS), SQL, A/B Testing, Data Analytics, Slack, Git, Scientific Data Analysis, Science, Data Visualization, Natural Language Processing (NLP), Management, Project Design, Fraud Prevention, ETL, PostgreSQL, Artificial Intelligence (AI), Time Series, Cloud, Statistical Modeling, Data, Analytics, Quantitative Finance, Quantitative Risk AnalysisDecision Sciences Chapter Lead and Senior Manager
2019 - 2020iFood- Led the anti-fraud department's data team during a chargeback crisis and reduced the chargeback by 90%, from over 3% to less than 30 basis points.
- Proposed and assembled three new horizontal structures within the data domain, changing the structure from a single back office unit to squads across domains, which impacted over 80 employees.
- Designed a new automated evaluation process based on graphs and reduced the performance evaluation time from two weeks to two days.
- Made an FP&A cohort model, which resulted in the company changing how it spent a monthly $10 million marketing budget.
Technologies: Linux, Machine Learning, Data Visualization, Project Design, Data Analysis, Fraud Prevention, Logistics, Growth, Financial Planning & Analysis (FP&A), Python 3, Python, Data Science, Time Series Analysis, Amazon Web Services (AWS), SQL, A/B Testing, Data Analytics, Slack, Git, Scientific Data Analysis, Science, Natural Language Processing (NLP), Management, Technical Hiring, Hiring, Graphs, ETL, Databricks, PostgreSQL, Artificial Intelligence (AI), Time Series, Cloud, Data Engineering, Statistical Modeling, Data, Analytics, Quantitative Finance, Quantitative Risk AnalysisData Specialist
2017 - 2018Foxbit- Recovered crucial customer data lost during a third-party platform provider (a white label solution) crash. This drastically mitigated any effects the crash may have caused on the company's customer base.
- Detected a product flaw that allowed the customer to exploit its credit system. Over 100k USD and 30 BTC were recovered from customers abusing that product flaw.
- Automated the financial reconciliation process the company had.
Technologies: Python 3, Amazon Web Services (AWS), Crypto, Data Analysis, Linux, Slack, Git, Scientific Data Analysis, Science, Data Science, Data Visualization, Natural Language Processing (NLP), Machine Learning, Project Design, ETL, Time Series Analysis, SQL, A/B Testing, Data Modeling, Time Series, Data Mining, Cloud, Data Engineering, Statistical Modeling, Azure, Data, Analytics, Quantitative Finance, Quantitative Risk AnalysisLead Data Scientist
2016 - 2017Mira Educacao- Hired and structured a small team consisting of four data scientists and engineers.
- Oversaw data engineering and data analytics projects and processes.
- Built an Optical Mark Reader for the mobile app, which could scan and correct test sheets.
Technologies: Amazon Web Services (AWS), Python 3, ETL, Data Analysis, Machine Learning, Computer Vision, Data Visualization, Linux, Slack, Git, Scientific Data Analysis, Science, Data Science, Natural Language Processing (NLP), Management, Project Design, Technical Hiring, Hiring, Time Series Analysis, SQL, A/B Testing, PostgreSQL, Data Modeling, Artificial Intelligence (AI), Time Series, Data Mining, Cloud, Data Engineering, Machine Vision, Statistical Modeling, Data, AnalyticsBusiness Analyst
2016 - 2017Nubank- Detected an issue in the company's underwriting model in which it misrepresented the operational cost of customers. As a result, the company was onboarding unprofitable customers, severely impacting its P&L.
- Defined main KPIs for the new customer credit policies and built the monitoring tools and processes for those.
- Perceived an issue with how the company grew its operational team and offered an alternative that reduced the customer experience cost by 60%.
Technologies: Python 3, Databricks, Linux, Amazon Web Services (AWS), MongoDB, PostgreSQL, SQL, Data Visualization, Data Analysis, Slack, Git, Scientific Data Analysis, Science, Data Science, Natural Language Processing (NLP), Project Design, ETL, Time Series Analysis, A/B Testing, Time Series, Cloud, Statistical Modeling, Data, Analytics, Quantitative Finance, Quantitative Risk Analysis
Eloi Ugo Pattaro
Data Scientist and Developer in Berlin, Germany
Member since August 1, 2022
Eloi is a data scientist and manager with a bachelor's degree in physics and over seven years of experience working for market-leading tech startups such as Klarna and iFood. Mainly working with Python, Eloi focuses on machine learning, data science, computer vision, and complex algorithms. Eloi has worked on key anti-fraud, growth, marketing, logistics, and HR projects and has managed over 30 professionals, including tech leads and data scientists, analysts, and engineers.
Eloi is now available for hire
Experience
- Data Visualization 11 years
- Data Science 11 years
- Python 3 7 years
- Project Design 7 years
- Machine Learning 7 years
- Management 5 years
- Fraud Prevention 3 years
Availability
Part-time
Preferred Environment
Linux, Amazon Web Services (AWS), Python 3, Git, Slack
The most amazing...
...achievement I've had is updating a fraud prevention system, significantly increasing its performance, and reducing the fraud rate from 3.5% to less than 0.3%.
Employment
Experience
- iFood – Performance Evaluation Graph System
Created a graph-based system that decides who should evaluate the performance of other employees at the end of a cycle, enabling the company to automate a large part of its evaluation process and significantly increase efficiency.
- iFood – Anti-fraud Decision Engine Refactoring
Collaborated with my team to reduce the complexity of a fraud prevention system, for which the company had a convoluted undocumented system that wasn't working as intended, and nobody knew why. We reduced over 500 policies to less than 20 while dramatically increasing the system's performance and lowering the 3.5% fraud rate to less than 0.3%.
Skills
Languages
Python 3, Python, SQLParadigms
Data Science, Management, ETLPlatforms
Amazon Web Services (AWS), Linux, Databricks, AzureOther
Scientific Data Analysis, Science, Coding, Data Visualization, Machine Learning, Data Analysis, Project Design, Fraud Prevention, Technical Hiring, Hiring, Time Series Analysis, A/B Testing, Data Analytics, Data Modeling, Artificial Intelligence (AI), Time Series, Data Mining, Cloud, Data Engineering, Statistical Modeling, Data, Analytics, Quantitative Finance, Quantitative Risk Analysis, Physics, Natural Language Processing (NLP), Computer Vision, Logistics, Growth, Financial Planning & Analysis (FP&A), Graphs, Crypto, Machine VisionTools
Git, SlackStorage
MongoDB, PostgreSQL
Education
- Bachelor's Degree in Physics2011 - 2016Universidade de Sao Paulo - Sao Paulo, Brazil
