Being promoted to solution architect allowed me to directly work with other teams and have a greater vision of our business model, helping design solutions that would have long-term impact on our users.

As Tech Lead I’ve been in contact with stakeholders and managers aligning dates and ensuring the delivery of the platform I’ve worked on to build. I was chosen as Tech Lead due to my extensive knowledge of the tool and rigorous deliveries. I’ve managed to improve the overall satisfaction of our product by employing a better strategy to our deployments and integration between our services that led to smoother deliveries. I was also hands-on on the product which made everything more natural when discussing timelines and tasks with my subordinates.

While working at SAP, I was also working at TeiaLabs, mainly developing API's that would be used to serve in-house models and models from external providers with enterprise level scalability. At TeiaLabs I did a more software engineer focused work, better designing the architecture of the system, designing the test cases for the tools, configuring CI/CD for all of our API's and setting up the infrastructure for the code to run, primarily using AWS.

After being promoted, I spearheaded an initiative to improve our quality process by redesigning our validation system that was outdated and hard to maintain. It used Jenkins to automate the testing and reporting of the issues found. I re-wrote all the communication protocol with our system and designed the architecture to easily add new tests and execute them in parallel, speeding up the time taken to run them and improving the maintainability of the code.

At SAP I was working with a legacy system in the team of Data Services. There I started by giving support and implementing new features to the system that included new external connectors to databases and external systems that clients would like to consume their data from and working on performance critical modules of the system, where I managed to improve by 50% the speed at which we were able to process batch events. This was an Agile team where each sprint would be defined by features and client tickets that developers would need to take on. It was my first direct contact with clients where I could understand their needs, understand the application and provide a solution that would resolve their problem.

As an intern at TeiaLabs, I was working with image recognition. At first I was tasked with working with OCR for processing documents that should be digitized, I took an approach to first finetune the model to better understand the documents we were processing, so I had to annotate and augment the data we had to improve the results we were having. It was at this moment that I did Andrew Ng's course to understand and dive deeper into machine learning and artificial intelligence in general.

As an AI researcher at a laboratory in PUCRS, I developed a flood simulation system that was capable of receiving events from multiple intelligent agents, updating the internal system structure with these agents' actions and generating the new state visible for the agents. The purpose of the simulation was to validate how well these agents would perform in a real life scenario where there were victims and resources that they would need to manage. I also developed one agent based on hierarchical reinforcement learning that would be used as a baseline for my supervisor and doctorate to compare his solution, which was based on YOLO with Deep Q Learning.

Worked as one of the first developers in the company performing activitiessuch as: task automation, documentation and construction of projects and dataprocessing, particularly with applications involving Office 365 APIs, reports ofdata consumption with Power BI and registration automation systems.