Senior Data Scientist - Data & AI

Our Technology area is a fundamental pillar of our business, responsible for our digital transformation. Here, we create technological solutions by always conceptualizing and developing automation techniques and processes, whether to improve our teams' work experience or increase the delivery capacity of our products and services.

We are responsible for a gigantic infrastructure that serves Globo's main video products, such as Globoplay, Premiere, and the Globo Channels. Additionally, we are responsible for the live broadcast of major events - from state football championships, the World Cup, Olympics, Rock in Rio, and even the live cameras of Big Brother Brasil, which is available 24x7 for all Globoplay subscribers.

The Data & AI front develops data platforms, experimentation, machine learning, recommendation, and audience segmentation for Globo's digital products. These platforms impact millions of users per month and support AI models applied to various challenges, such as content personalization, contextual advertising, churn prediction, fraud prevention, inventory optimization, and much more. We are looking for a Senior Data Scientist to work in the Anti-fraud front, with solid experience in Machine Learning applied to fraud detection and prevention, autonomy to conduct end-to-end projects, and the ability to translate data into scalable risk decisions. Generative AI tools may be used as support, but the daily focus is ML.

We create environments where all people are included, welcomed, and valued. We strive to be an increasingly diverse Globo, in representation and thought, learning from each other and stimulating creativity and innovation.

Work model: Hybrid in locations with a Globo office (RJ, SP, POA, Brasília, Recife, or BH)

Responsibilities and assignments

How your day-to-day will be:

• Build, train, and productize ML models for fraud detection and prevention (classification, anomaly detection), dealing with highly imbalanced datasets;
• Develop and maintain real-time (or near-real-time) scoring pipelines for in-flow risk decisions;
• Monitor performance and drift of models in production, with continuous retraining — essential in an adversarial scenario where fraud patterns change constantly;
• Perform feature engineering on transactional and behavioral data to capture fraud signals;
• Collaborate with risk, product, and engineering teams to balance fraud blocking and legitimate user experience (tradeoff between false positives and false negatives);
• Explore the use of generative AI as a support tool (acceleration of analyses, prototyping, documentation), when it makes sense;
• Act as a technical reference in the team, supporting architecture decisions and contributing to the development of more junior people;
• Work in a multidisciplinary team, collaborating with ML engineers, data engineers, and POs to create data-driven solutions.

Requirements and qualifications

What you need:

• Completed higher education in Data Science, Statistics, Mathematics, Computer Science, or related fields;
• Solid experience as a Data Scientist, with autonomous performance in end-to-end ML projects (from conception to production deploy);
• Proficiency in Python and SQL for data analysis and modeling;
• Experience with classification and/or anomaly detection, including handling imbalanced classes;
• Familiarity with metrics suitable for imbalanced scenarios (precision/recall, PR-AUC, business cost-oriented confusion matrix);
• Knowledge of Machine Learning frameworks (Scikit-Learn, XGBoost/LightGBM; TensorFlow or PyTorch are a plus);
• Familiarity with MLOps — model monitoring, drift detection, and continuous training;
• Experience in cloud computing (preferably GCP);
• Experience with handling large volumes of data;
• Knowledge of software development best practices, CI/CD, and versioning (Git);
• Good communication and teamwork skills.

Knowledge that sets you apart:

• Previous experience with anti-fraud, risk, or anomaly detection in production;
• Graph-based modeling for network fraud detection (gangs, fraud rings);
• Experience with low-latency online prediction APIs;
• Model explainability (SHAP, feature importance) to justify blocking decisions;
• Familiarity with generative AI / LLMs applied to productivity or analysis;
• Advanced experience with MLOps in GCP (Vertex Pipelines, Model Monitoring);
• Intermediate/advanced English (writing, reading, and conversation).

Additional information

Why come to Globo: Come and let's create stories and technologies together. At Globo, the meeting of stories with technology moves a diverse team passionate about innovation. We have the chance to bring Brazilians closer, creating new ways to connect with them. Here, everyone makes the transformation happen.

Our selection processes are happening 100% remotely. #vempraglobo

Want to know more about what it's like to work at Globo? Come follow our daily life, challenges, and achievements on our networks: https://www.linkedin.com/company/globo https://www.instagram.com/vempraglobo https://www.youtube.com/c/VemPraGlobo/

Your personal data will be used for candidate selection purposes to fill open positions at Globo. Personal data collected during the selection process of candidates who are not admitted at this time may be kept in databases for consideration in future selection processes. All collected data will be treated according to strict information security standards and in full respect for applicable privacy laws, including Law 13.709/18, and in accordance with our privacy policy, available at the link: https://vempraglobo.g.globo/files/politica_de_privacidade.pdf

By proceeding, you are aware and agree to carry out your application under these terms.