- I Developed integrations for complex banking systems, with systems communicating asynchronously using RabbitMQ. The systems are developed in GO, C#.NET Core and Python. Some of these applications use Computer Vision to analyze documents, fraud and biometrics, using OpenCV, Keras and Pytorch, with supervised and unsupervised models.

- I Designed, created and developed C++ and C Drivers and APIs for devices used in bank ATMs. I also developed software for analyzing money bills using computer vision to prevent fraud in ATM deposits.- Designed and developed software for bank treasury systems, using Rest API developed in GO, using SOLID and DDD architecture, with access to SQL Server database hosted on AWS. The big difference presented in this application was the use of equipment for recognizing banknotes for money deposited and withdrawn at ATMs in bank treasuries, almost completely reducing all losses arising from human errors.

- Implementei API REST para plataforma de emprestimos consignados bancários em .Net Core utilizando SQL Server com Views e Stored Procedures sincronizando dados com correspondentes bancários. Este processo possibilitou o aumento do número de propostas de emprestimos vindo de correspondentes bancários, e também possibilitou com que emprestimos fossem rapidamente efetivados.- Implementei e fui responsável pelo projeto do ECO (Emprestimo Consignado OnLine), que consiste na comunicação com o DataPrev (orgão do governo responsável por aposentadorias e pensões) para buscar as informações e limites de crédito dos clientes, e também toda efetivação do emprestimo com bloqueio de margem consignada diretamente na DataPrev. Neste projeto, foi utilizado GO para implementação da API REST das rotinas de processamento de concorrencias de transações e também um serviço windows desenvolvido em C# para orquestração entre diversos sistemas que se utilizam do sistema ECO.- Projetei e implementei um sistema em python, para análise automática de documentos utilizando Visão Computacional através da utilização de OpenCV e Keras. Este sistema reconhecia o tipo de documento e também faz o reconhecimento dos caracteres (OCR), retornando dados precisos de cada documento. O sistema foi gerado em Python com framework Django para ter um CRUD de manutenção de registros e parametrização do sistema. Este projeto possibilitou a redução de analistas de documentos, diminuindo 95% dos funcionários que tinham essa função, restando só aqueles que devem analisar documentos muito antigos e razurados que o sistema tinha dificuldade de reconhecer.

- Developed several embedded systems to automate agricultural machines in GO, C#, C++ and Python.- The main systems developed will use computer vision in fruit sorting equipment, where they separate fruits by size, color and defects. These systems provide the customer with the standardization of their products, guaranteeing a profitability gain of more than 25% compared to the traditional classification model.- Developed several REST APIs in GO, C# and Python with the publication of these services on AWS.

- I was part of the game developer's tools team, where I developed several artificial and physical intelligence tools for online game developers to build their applications. The development was done in C# with Unity, using the SQL Server database.- Developed routines in C# for the MMO Taikodom Live Universe.- I developed the Bitverse Ads tool, which consists of software that implements advertising in online games.

- Developed application in C# Web Forms and WebServices for tax and financial auditing. This application replaced the company's previous version, allowing greater reliability and generating audits remotely, thus guaranteeing a great performance gain for the audit team.- Developed audit routines in Store Procedures and Views in SQL Server, generating rapid external audits when necessary to search for information that the systems did not generate in their reports, thus ensuring the delivery of audits in a timely manner.

- Developed computer vision projects and systems through image processing algorithms and artificial neural networks in C++ and C#.- Development of an MLP Artificial Neural Network for use in various company vision systems.- Development of an image processing library in C#, replacing the version developed in C++ and adding more functions. This process allowed for faster development of the development team, allowing the team to receive between 30% and 40% more projects for development at the same time, thus ensuring better performance and greater profitability for the company.- Development of people flow counting software for a shopping center using C# language with artificial neural networks. This product entered the market, allowing the company to increase its revenue by 15%.- Development of inspection and verification applications through computer vision, for analysis of electronic components, automotive parts, cookies, GM and FIAT vehicle assemblies.- Development of an LPR (License Plate Recognition) application for Brazilian vehicles, using this application on the border between Brazil and its neighboring countries for police and tax control and inspection. Another use of the application was in highway toll booths, where each ticket is marked with a specific vehicle plate, thus guaranteeing toll booth services in case of problems. This application was also widely used to access residential and commercial condominiums and shopping centers, ensuring that only those who have a license plate ticket can leave the condominium, preventing fraud and vehicle theft.

- Development of applications in Visual Basic, C# and SQL Server database.- Development of several modules of the company's ERP, including accounting modules, billing, tax books, POS, concentrator and accounts payable and receivable modules.- I was responsible for approving the integration of systems with TEF (Electronic Funds Transfer) with credit and debit card administrators (Visanet [current Cielo], Redecard, Tecban, Amex [American Express] and Banrisul).