Ellton Barros

Results-oriented Data Scientist Specialist I (Tech Lead) with a strong background in making technical decisions, analyzing risks, and building advanced machine/deep learning models. Recognized as a technical reference within data science teams, specializing in Natural Language Processing (NLP) and Deep Neural Networks (Bert, RoBERTa, etc.). Highly skilled in reproducing state-of-the-art results to solve complex NLP problems. I also received the Gold Astronaut Award for outstanding results with NLP.

I worked as a Tech Lead and Data Scientist in the PCD (Cognitive Document Processing) project, developing an intelligent architecture for document classification and information extraction using computer vision, OCR, deep learning, and AI. Additionally, as a Machine Learning Engineer, I created scalable AI services through REST APIs for the ICN project. I built and monitored Chatbots using NLP and received a banking innovation technology award for developing a billing chatbot via Whatsapp. In the Tech Lead role, I contributed to the Positive Biometrics project, utilizing facial recognition for fraud detection in retirement plans. I modernized ML project architecture through MLOps implementation with MLflow. As a Data Scientist, I created an ML model for ATM failure prediction nationwide, optimizing maintenance and minimizing unavailability. I also developed an ML model to identify high-risk security incidents in bank branches, reducing risks such as robbery, theft, and kidnapping.

In this position, I was acting as Angular 6 (front-end) and Java (back-end) developer.