José Henrique Blanco Email and Phone Number

Ravelin é uma startup que utiliza machine learning e network analysis para prover detecçãode fraude em tempo real para marketplaces. Como parte do time de detecção eu:- Desenvolvi modelos supervisionados de detecção de fraude do tipo card not present,incluindo a identificação de padrões de fraude, extração de features e monitoramento deperformance. Já implementei modelos para mais de 10 clientes em diferentes segmentosde mercado.- Construí e aprimorei de pipelines para o processamento de dados e treinamento demodelos. Algumas das melhorias incluem a criação de um sistema de monitoramentos deuso de recursos, melhorias no sistema de logging e a redução do tempo de execução econsumo de memória.- Reestruturei o ambiente de desenvolvimento da equipe de forma a simplificar os processosde exploração de dados, desenvolvimento de modelos e compartilhamento de informações.- Utilizei redes neurais para modelar padrões comportamentais de usuários para utilizá-los na detecção de fraude.- Servi como tutor para os membros junior da equipe e integrantes de times não técnicos.- Atuei como especialista técnico do time, atuando na resolução de problemas no ambientede produção.

As a data scientist working for The Quarry, I have the opportunity to work in a multi-disciplinary team to develop end-to-end solutions that make use of the whole data science stack. Some of the tasks that I have undertaken were:• Structuring approaches to tackle data-intensive projects, starting with the data sources all the way to modeling.• Fetching, wrangling and cleaning data using SQL, Python, Pandas and other open source tools.• Drawing insight from data through visual analysis, summary statistics, hypothesis tests and machine learning.• Developing, validating and deploying predictive models.• Presenting findings in a clear and concise manner through reports and presentations.

After finishing the data science immersive course myself, I was invited to join the instructional. Among my assignments are:• Aiding the students to develop an intuition on the abstract concepts of algorithms and machine learning models.• Guiding the students through the data science methodology, indicating the importance of each step and how to avoid common pitfalls, as well as what are the best practices when it comes to data preparation, training and model deployment.• Providing students with feedback on their respective projects, proposing different approaches to data visualisation, preprocessing, feature selection and model tuning.

The data science immersive course was truly an intense and rewarding experience, during which I have:• Got acquainted with data handling, exploration and transformation, as well as machine learning and network analysis using Python and open source tools.• Enhanced my teamwork and presentation skills through group projects.• Worked on projects that simulated real-world scenarios, covering topics such as web scraping, clustering, supervised learning and recommendation systems.• As my capstone project, I have applied Natural Language Processing techniques to develop a model capable of assign rating to reviews extracted from Amazon, as well as detect abstract concepts such as the usefulness of a document.

During the internship at Petrobras I was part of projects that ranged from power generation and energy efficiency to costs optimization. Besides the engineering aspects of such endeavors, some of the tasks that I have undertaken were:• Using the company’s database and external data sources to assess the energy market and aid in decision making.• Enhancing the department’s data pipeline and developed a simple dashboard, which led to a reduction in the data latency and finer control over the department’s KPIs.• Detecting frauds and faults in the department’s numerous power plants using time series analysis.• Proposing new products based on the ever-changing energy market regulation.