About Marcos Eduardo
Data platforms are my day-to-day
I work as a Junior Data Platform Engineer at MadeiraMadeira, where my focus is data infrastructure, cloud automation, and tools for technical teams.
I work with Python, SQL, and Google Cloud, including tools such as BigQuery, Airflow / Cloud Composer, Dataform, and Dataflow. My projects also give me room to explore AWS and Docker.
Learning what happens underneath
Rust is my current study focus. I’m working through ownership, borrowing, types, and error handling, while spending more time on memory management, concurrency, and operating systems.
I’m interested in networking and streaming: the protocols and systems that let applications exchange data and deliver audio and video. Longer term, I want to explore systems software with Rust and C++, and learn more about embedded systems.
Building with data and AI
GCP AI Copilot is my exploration of documentation retrieval: collecting GCP documentation, generating local embeddings, retrieving relevant context with Qdrant, and exposing the result through FastAPI and a React interface. The MCP integration remains a planned extension.
TwitchPipeline brings together another set of interests: stream metadata, PySpark, AWS orchestration, and contextual analysis with Gemini. My experience also includes LLM-assisted code review.
A little more Ryota
I’m based in Curitiba, Brazil, and study Computer Science at UNINTER. Away from work and study, you might find me playing Final Fantasy XIV, watching Monogatari, listening to music, or experimenting with Linux.
Games, music, and streaming are part of what makes systems programming interesting to me. I want to understand the software behind the experiences I enjoy, and build tools I would use myself.
What this site is for
This is where I keep project write-ups, engineering notes, and things I learn along the way. The reports are a record of that process, including older studies in cloud and data engineering.
The site itself is also a project: React and Vite, Bun for the development workflow, and static hosting on Cloudflare. I’m keeping it small enough to understand and maintain myself.
Working on data platforms, systems, or something related? I’m happy to exchange ideas. Find me on LinkedIn.