Thiago Henrique Coraini

This project was done while I was working on my master thesis. It was a fork from the Akka project; a Scala framework heavily focused on the Actors model. As part of my work, I added two capabilities to the existing framework: actor automatic distribution and migration.

The idea was to make developing distributed applications easier and seamless (with automatic distribution) and more fault-tolerant (with actor migration).

The complete thesis, in Portuguese, can be found at the link below: • https://teses.usp.br/teses/disponiveis/45/45134/tde-18032012-134507/publico/DissertacaoThiagoCoraini.pdf

While working on a system that would calculate and display an optimized way to load products in a container (e.g., a truck), we faced difficulty meeting customers' requirements who had particular loading patterns for some of their goods.

I was the proposer (and for some time the main implementer) of several changes in our current algorithms (which were based, mostly on academic research) that would satisfy the customer loading patterns while also attending to all the other system requirements like optimizing the load and meeting hard restrictions (e.g., max stacking weight).

Working together with the data science team in our squad, I was in charge of delivering the API that would expose, to other services in our company, machine learning models and insights for our real estate customers. Those included, for example, information about pricing ranges and comparisons with similar properties.

I was the main person responsible for rolling out the microservice that would deliver that information through RESTful APIs. At the same time, I coordinated with the DS team to make the data handover reliably flow from the models to our web application.