About

Aug 8, 2026 · 3 min read

Felipe Cardoso

I’m Felipe Cardoso, a Brazilian software engineer. I live in Tokyo and have been working on backend, cloud infrastructure and distributed systems since 2018, along a path that runs through a bank, a consultancy, a global brewer and, now, an electric mobility startup.

I like owning a whole system: understanding the problem, deciding the architecture, writing the code, shipping it and answering for it when it breaks. Language, database and cloud choices come after the problem, never before.

Why this blog exists

To share what I learn and what I think about software engineering and technology. The technical posts come out of what I am actually building, including the parts that went wrong and rarely show up when only the final version gets told. The rest is opinion: how we build software, what is worth doing and what is just fashion.

I write in Portuguese, English and Japanese. Comments are anonymous by default, because disagreeing with me should not require an account anywhere.

The opinions published here are my own. They do not represent or necessarily reflect the positions of the companies I work for or have worked for, or those of their customers.

What I am doing now

I’m a backend and infrastructure engineer at eMotion Fleet, a Tokyo startup building fleet and energy management software for electric bus operators in Japan, including Nishitetsu and Nagasaki Bus.

The heart of the product is a real-time decision problem: how much power to send to each bus plugged in at the depot, without blowing past the contracted energy limit, while making sure every vehicle leaves on schedule with enough charge. In my first month I proposed an AI model for that problem and built the prototype that became the platform’s core decision engine, the same engine that helped the company pass its first customer acceptance test and reach its first production go-live in September 2025.

Beyond the engine, I look after the AWS platform, lead product streams from design to delivery and handle production incidents.

Where I have been

AB InBev, 2022 to 2024. I worked on BEES, the B2B commerce platform with over 1.5 million users and $49 billion moving through it every year. I built microservices for the rewards program, which reached 2.3 million members with an NPS of +64, and the reusable Terraform and Terragrunt infrastructure behind the company-wide migration from AWS to Azure. On the eve of the South Africa launch, I fixed the production issues that were threatening the date.

Itaú Unibanco, 2019 to 2022. I finished there as tech lead on a pricing platform, leading three developers: corporate price negotiations that used to depend on spreadsheets and take hours dropped to minutes, with the calculations and client-level profitability analysis automated in C# and ASP.NET Core and the interface built in Angular. Before that, in fraud prevention, I mined historical patterns with Python, SAS and Power BI and implemented T-SQL detection rules that helped cut debit-card fraud by about 12%.

SAS, 2018 to 2019. Data engineering consulting: anti-corruption corporate-card analytics for Vale and nationwide fuel supply and demand forecasting for Raízen’s station network.

Education

  • PUC Minas, Postgraduate Degree in Distributed Software Architecture (2024-2025)
  • Universidade Federal do ABC, BTech in Computer Science (2018-2023)
  • Universidade Federal do ABC, BS in Science and Technology (2018-2023)

Contact

I’m open to remote roles with globally distributed teams.