● Trained computer vision models using YoloV5 and optimized them for the company's embedded systems using NVidia's TensorRT tool.● Used tools such as Memray, mprof, cProfile, and pyprof2calltree to generate memory and CPU usage diagnostics, and produced performance reports.● Conducted performance troubleshooting and optimized Python code, reducing memory usage by 75% and decreasing production incidents.● Developed a Proof of Concept (POC) using React and Node for a 3D product visualization platform on the web.● Implemented a POC to enable web-based viewing of images transmitted by a drone in a monitoring project, creating a viewing page with React and an HLS stream using Node from the drone's RSTP stream.

● Implemented from scratch a client-server application for data visualizationin the Search team using Angular for frontend and C++ for backend.● Integrated to this application a widely used enterprise plug-in thataggregated value to data visualization, simplified testing, and improvedoverall maintainability.● Followed best practices for code versioning using Fig.

● Structured a generic vehicle customization system in C#/Unity andintegrated it with the infrastructure of PKXD, a game played by more than20 million players around the world.● Assisted in the development of a grid-based showcase system,mathematically modeling rotations and translations using Linear Algebraand Computer Graphics theory.

Worked as a coordinator helping to manage more than 10 projects by semester, guaranteeing its success and bringing ideas and solutions to problems in the group's development processes.

I developed games (alone and as a team) and participated in several Game Jams. I learned Unity, C# and became familiar with Design Patterns and Software Architecture, subjects in which I gained a deep interest. My projects can be found here: https://jpass.itch.io/

Use of computer vision to automate processes in the soybean research area, facilitating seedling measurement, classification and vigor assessment, directly impacting the targeting of seeds for human consumption or planting. Using OpenCV, Python and Deep Leaning techniques for image processing. This project was funded by São Paulo State Research Support Foundation (FAPESP).

I created tools to help the automatic customization of multivideo summaries using Recommendation Systems concepts. Made with Python, Django and PostgreSQL. Funded by the National Council for Scientific and Technological Development (CNPQ).