Looking for a job

Felipe Daiha

@felipedaiha

Message

Data Scientist at Mercado Libre optimizing ad bidding to cut CPM by 10%.

Brazil

Message

At Mercado Libre, I train and deploy machine learning models and bidding optimizers for programmatic advertising, helping maximize advertiser ROI and cut CPM by 10%. I also build data pipelines that supply model training and analytical dashboards, accelerating execution times by 50%.

I designed ML engineering applications at Mercado Libre that reduced internal server errors by over 70%, and created an observability and alerting system adopted by the team. Before that, at Mutual, I developed credit-risk models that supported a loan portfolio expansion of 20M+ BRL while controlling 90-day default rates to 3%.

At Mutual, I also implemented AWS data pipelines that reduced costs by 90% compared with the legacy system. My work spans data science, data engineering, and MLOps, including Python and R modeling, cloud platforms, and analytics.

Experience

Work history, roles, and key accomplishments

Current

Data Scientist

Current

Mercado Libre

Apr 2023 - Present (3 years 7 months)

• Trained and deployed Machine Learning models (Regressors, Recommenders / Similarity Search) and optimizers (Functions minimization) for programmatic advertising bidding components to maximize advertisers ROI by slashing CPM by 10% and achieving target values for Awareness, Reach, CPC, CPA and ROAS.

• Implemented scalable data engineering (ETL) pipelines utilizing Dataflow, BigQuery, Python and

Data Science & Data Analysis Instructor

TripleTen LatAm

Feb 2025 - May 2026 (1 year 3 months)

• Conducted 200+ synchronous Q&A sessions to resolve student queries on active exercises and projects, maintaining a near-perfect 4.95/5.00 average CSAT and developing complementary platform materials to reinforce learning from foundational to advanced concepts for aspiring Data Scientists and Data Analysts.

• Mentored students in 1x1 sessions, providing hands-on guidance in practical technical s

Data Scientist

Mutual

Dec 2021 - Feb 2023 (1 year 2 months)

• Trained and deployed Machine Learning models (Ensemble models, Clustering) for credit risk scoring solutions, which allowed to expand the loan portfolio by 20M+ BRL, controlling 90-day default rates (OVER90) to 3%, as well as behavioral modeling for payment and renegotiation propensities, achieving a renegotiation rate of 40%.

• Architected and implemented an end-to-end data engineering infrast

Data Science Intern

Mutual

Mar 2021 - Dec 2021 (9 months)

• Integrated Business Intelligence tools such as Power BI, Metabase, and Google Data Studio (Looker) to seamlessly query and visualize data from PostgreSQL and MongoDB databases for multiple teams in the company.

Data Science Intern

LEMETRO - Laboratório de Experimentos em Mecânica e Tecnologia de Roch

Aug 2018 - Aug 2020 (2 years)

• Simulated over 1000 ballistic scenarios per terrain configuration utilizing computational modeling (RocFall v4.0) to rigorously analyze physical parameters, including restitution coefficients, block masses, and velocities, storing these informations in MySQL database.

• Evaluated the critical impact of angular velocity and terrain breaks on rock block reach using R programming language for data

Junior Database Administrator

LIEG - Laboratório Interdisciplinar de Estudos Geoambientais

Mar 2019 - May 2019 (2 months)

• Managed and processed extensive fieldwork datasets from the 2011 landslides in the Rio de Janeiro mountainous region (Nova Friburgo and Itaoca) utilizing MySQL database, ensuring high data quality, structured storage and efficient querying of complex environmental records to facilitate accurate historical analysis and reporting for research proposal.