About Myself

Data Scientist graduated as a Computer Engineer from the University of São Paulo, I have experience with prediction models for different tasks such as investment allocation, pricing, assortment, and demand forecasting. I’m currently enrolled in the Masters in Computer Science program at the University of São Paulo, researching with a focus on improving the predictive power and scalability of graph neural networks for high-dimensional graphs. I also have experience with Machine Learning code productization, machine learning engineering, and data science team management.

My interest areas include complex networks, deep learning, machine learning engineering, and data products.

I also am the creator and maintainer of the scikit-net, an open-source library with machine learning algorithms for complex networks.

Career

Stone Payments

I’ve worked at Stone Payments as a summer internet with a routing system for the sales team. This system generated optimized sequence of visits aiming to minimize the traveled distance while maximizing the expected LTV of the visited leads. This system was integrated into a more general commercial intelligence system at the company.

Preliminary results generated an estimate of over R$ 10MM in productivity gains for the company in the year of 2019.

Used frameworks and tools: Python, Salesforce, Complex Networks

Big Data

I’ve worked at Big Data for over three years, from intern until a full data scientist with team management responsibilites. My main projects and accomplishiments at the company were:

Worked as head of the assortment product, being responsible for executing, plan and propose methodological improvements to the product. Mentored the development of 3 data science squads in order to align their work with the product demands.
Managed a team of data scientists in several retail projects.
Technical and methodological refactor of investment allocation model to improve its performance, quality and effectiveness.
Worked on the productization of all of the company’s products, allowing them to be reusable. Also improved their performance.
Fully developed a go-to-market strategy for a beverage company, comprising marketing allocation, assortment, pricing points and demand predictions per region.
Developed models for forecasting demand, pricing and assortment.
Developed quality assurance pipelines for machine learning data pipelines.
Developed an experimentation framework based on multi-armed-bandits strategy for pricing optimization.

BTG Pactual

At BTG Pactual my main job is related to risk reduction at the retail banking. My developed projects involved:

Bad behavior for credit users detection
Self-financing and money laundring detection
Fraud detection and mitigation at transaction time
Fraud dectection and mitigation at onboarding time

Portfolio

Data Science Teaching Production

Medium Posts

I have several educational posts on Medium, mainly focused on implementing solutions and papers from scratch to assist my readers into understanding the inner workings and complexities of what is being presented. You can access my profile here. Here are my most recent posts:

LinkedIn Profile

In my LinkedIn profile I post regular tips about Data Science, Research and general productivity to help my readers become better professionals. You can check this out at my profile: https://www.linkedin.com/in/tiagotoledojr.

Open-Source Development

sknet

EPCSAC

Publications

Toledo, T., (2021). sknet: A Python framework for Machine Learning in Complex Networks. Journal of Open Source Software, 6(68), 3864, https://doi.org/10.21105/joss.03864

T. Toledo Junior and S. Bruschi. “EPCSAC - Extensible Platform for Cloud Scheduling Algorithm Comparison”, in Anais Estendidos do XXI Simpósio em Sistemas Computacionais de Alto Desempenho, Evento Online, 2020, pp. 46-53, doi: https://doi.org/10.5753/wscad_estendido.2020.14088.

Tiago Toledo Jr.