Hello, I'm Romulo Gatto

As my first delivery at Tenchi, I handled the lambdas and batch job sizes better, given that only some of the step function runnings needed the most significant size, and most of our running had an idleness of more than 50%. After accomplishing that, I started to apply some clean code rules and refactors. After that, I got back into the costs, now saving on the DB individually for each of the runnings usages. Before starting each step function, I allocate only the necessary amount of ram/CPU to the steps.

I headed the migration and separation of a legacy PHP code used on WordPress into various small micro-services, occasionally hosted as Google Cloud Functions, this way we achieved better error handling and scalability in our operation. My main focus was on the development of the place app backend, helping the person on the pickup point side to have all the data required and deliver the items correctly. Besides that, one of the side goals was to remove all connections with WordPress and its Database, which involved the creation of new services and databases and applying the microservices concept to it. As other areas consumed data from WordPress, we gathered all the relevant info and developed a queue to keep the critical info on the old database up-to-date, without costing too much performance for any of the endpoints.

I started working with Java and Angular, with most of it being coded by an outsource company, while I managed all of it to work with the internal architecture using Git, Jenkins, and RPM. After it, the whole bank started to migrate into Agile, where I got arranged into a PDD squad (Bad Debt Provisioning), where we mainly worked to fulfill all the BACEN (Brazilian Central Bank) needs while upgrading all the back-level code on the way. There, I've mainly worked with COBOL and VB6, upgrading it to VB.net, and using SQL Server and Oracle databases to create stable software. Nowadays, I got a step down on the PDD lane, going to a squad that creates one of the biggest files that are needed for the PDD process. Here, we are currently finishing a project to evolve from a full COBOL process to only a small part in COBOL, plus Java, Python, ASP.net, and Cassandra. With this stack and a good architecture, we reduced from 10 to 5 days of processing.

Starting as an Intern, where I started working with some small local business software to start developing my communication and way of working. When I got promoted, I moved to the main team, assisting my senior partner in finishing a proprietary CRM software. After it was done, we had one of the biggest banks from Brazil as clients, where they needed to create a back-end of car insurance quotes, where they would send us a file generated on their Mainframe, and we would read it and do all the quotes in all the agreed insurances, then generate a batch file to send it back to them. All the works there were using Delphi (7, 2000, XE2, and XE3), plus Firebird and SQL Server.

I was mainly designated to generate reports for the superintendent, create VBA macros in Excel and Access to speed up the whole process and solve any hardware questions that were needed by the Tax Auditor.