Matias Zilli

Gestya is a satellite tracking and remote data management service designed to optimize the profitability of a vehicle fleet and improve the logistics. It's a last-generation system that, in addition to offering real-time location and vehicle status, provides exclusive access to MotorGuard monitoring and the tire pressure calibrator. The platform was originally running as a monolith using SQL, and I led a migration project to move the entire system to a microservice architecture using Kubernetes on top of AWS and MongoDB Atlas on cloud.

SmartPool is a comprehensive solution for monitoring and dosing chemicals in swimming pools in an optimal and automatic way, fully managed through an app. This was my own project, and it won an award in a Startup Weekend through Google Entrepreneurs' Day. I developed the app using AWS IoT Core, Lambda, and DynamoDB. The back-end consists of APIs that I developed using the Serverless Framework and a Raspberry Pi to sense the pool parameters.

In this project, we proposed a supervised deep learning-based model for accurate automatic cell nuclei segmentation. Given a colon tissue image, it begins with a deep convolutional neural network model to generate a probability map.

I outlined the full approach, containing the methods necessary to obtain a result by using a supervised deep learning method from an unannotated image dataset. I also demonstrated that the approach can perform on par or better than several state-of-the-art methods.

The project was supported by the European Commission under the Erasmus+ program in Budapest, Hungary. In 2017, it was recognized as one of the most important projects between Argentina and Hungary by the Faculty of Engineering and Water Sciences (FICH) at the National University of Litoral because it allowed both institutions to start a collaborative project with both countries.