Guilherme Amorim Avilla Gimenez

AI+Data & Software Engineer

Wander

Aug 2024 - Current (2 years 3 months)

At Wander I'm managing and building our internal Pricing Management System. We're innovating pricing by using AI solutions to get to the right price for both our hosts and our customers, focused especially on Interpretability, Customization, and Revenue Optimization. I'm also helping to maintain and evolve our internal Data Platform, by managing the Data Layer for pricing (by scanning and analyzing millions listings on OTAs and extracting insights for our operations).

Staff ML Engineer

iFood

Dec 2023 - Jul 2024 (8 months)

As a Staff Machine Learning Engineer, I have continued my responsibilities with innovation and collaboration. I've designed and set the ground for a "fully" Event-Driven ML Platform, focusing on reducing unnecessary costs during the execution of ML workloads but enabling different parts of the ecosystem to "react" to interesting and important events. For the MLOps CI/CD process, I've continued with the implementation of "modular" components which reacted to changes in the declarative file containing the ML Pipeline and its parts. Through a design of downstream pipeline (parent-child pattern) we were able to implement this design and reduce the overall build time of CI/CD pipeline for Machine Learning. Keeping up with the "pluggability" appr

Senior Machine Learning Engineer

iFood

Aug 2021 - Dec 2023 (2 years 5 months)

At iFood, I worked as a Machine Learning Operations (MLOps) Engineer. I lead the technical development and deliveries of the Machine Learning Platform products aiming specifically for extensibility and collaboration from the platform users (mostly Data Scientists and Data Analysts). This was achieved by using a pluggable infrastructure of the ML pipeline platform, integrated with other applications through a reactive infrastructure. I was also responsible for the unification of the platform API that was spread out in different legacy and un-maintainable applications to a simple API layer with a CLI. I've introduced a culture of quality with Code Review guidelines and testing standards to the team and also introduced RFCs and ADRs for asynch

Senior Machine Learning Engineer

Mercado Livre Brasil

Mar 2020 - Jul 2021 (1 year 5 months)

At Mercado Livre, I worked as an ML Engineer for Fraud Prevention. I've worked on designing a Data Science Workflow aimed at making our work repeatable, efficient, transparent, and correct. Also, having a 'formal' workflow helps us to understand the technical necessities for later creating or using applications to automate or keep track of specific tasks. I also helped in productionizing Data Science products focused on FP - given the challenges with imbalanced datasets, incorrect or indirect labels; this translates back to successfully monitoring model and business metrics, increasing data observability, and increasing model development, decreasing Data Scientists' burdens when creating new models.

Machine Learning Engineer

Santander Brasil

Apr 2019 - Mar 2020 (1 year)

At Santander I worked on building the Gutenberg Data Science Platform, a dedicated ML Platform to help scaling up the number of models in production. I made important contributions when creating the Batch Processing Engine and played an important role with deploying the first ML model in production (containing various integrations with Spark, CTRL+M, Kubernetes and Argo Workflows). I also contributed with our customized version of JupyterHub creating various components for Spawning Notebooks, dealing with LDAP Authentication and integrating various services with JupyterHub. Finally, one of my roles was to also help Gutenberg's Data Scientists to build better interfaces to preprocess data and also to train models, so I've built components to

Python Developer

DAFITI

Dec 2018 - Apr 2019 (5 months)

When working at Dafiti my responsibility was developing new features, and also maintaining our internal applications, to Pricing and Merchandising departments. On a daily basis, I had contact with Brazilian and LATAM (AR, AL and CL) analysts to understand better business needs and also provide further explanation of how our systems worked. I was responsible for the integration of internal systems for use in Latin America, automating different processes and workflows. I also contributed with the improvement of important Brazilian tools in our systems, always focusing in delivering value to the business and robustness to the system. We used Python (Django as our web framework) as main language to automate various internal business processes;

S

Software Developer

Jun 2018 - Nov 2018 (6 months)

At TDW Bi Consulting I worked on developing a Python platform distributed to Itaú-Unibanco data scientists based on a customized JupyterHub Application. In this project I had the opportunity to containerize notebooks using Docker and to use other tools like Jenkins, Puppet and Sonar to build a robust CI/CD pipeline.

Undergraduate Researcher

Instituto Federal de Educação, Ciência e Tecnologia de São Paulo - IFSP

May 2017 - Dec 2017 (8 months)

In 2017, I worked on a research, together with a team of doctors from the Medicine School of the University of Sao Paulo (FMUSP), called "Automation of the Coronary Artery Preparation Process for Atherosclerosis Studies". This research is directly involved with technological innovation because there is the development of two devices, a 'syringe pump' and a 'heat controller', that will control both process variables and will improve the procedure that is currently done manually.

Research Assistant

Mercer University

May 2016 - Jul 2016 (3 months)

Research: "Design and Development of Bio-Inspired Quadruped Robots" By adapting mechanisms from biological systems such as animals, engineering design has helped us to understand the related phenomena and the associated principles in order to develop novel devices and improve their functionality. Benefits from the bio- inspired engineering design and developments can be implemented in many endo- and exo-skeletal devices which are stronger and more stable. Students will design and develop bio-inspired quadruped robots which can walk, jog, and run using various technologies.

Guilherme Amorim Avilla Gimenez

About

Detail

Timeline

Résumé

Guilherme Amorim Avilla Gimenez

About

Detail

Behavioral traits