Antonio Artur de Holanda e Ayres de Moura

Polaris is an AI-powered cybersecurity solution that provides proactive, actionable insights by analyzing and filtering vast amounts of cyber threat intelligence data. It transforms fragmented information into tailored alerts relevant to an organization's specific attack surface, enhancing threat management efficiency and enabling security teams to focus on strategic actions. With real-time updates, custom notifications, and seamless integration with existing tools, Polaris streamlines cybersecurity operations and boosts overall productivity.

A platform that helps several companies tackle digital fraud. As a data science team member, I worked on several artificial intelligence models, such as phishing detection, website content classification, threat risk evaluation, topic models, and threat actor reputation assessment.

A platform for AI-assisted criminal investigations that is currently in the final development stages. As a data science researcher, I contributed to creating the platform's speaker recognition module.

An internal app used by Dell employees. I contributed to this project as an artificial intelligence researcher at Lead Dell, an institute financed by Dell in Fortaleza, Brazil. I created the app's speech recognition and natural language understanding models. Our team's end product was a module capable of recognizing more than 30 commands and executing actions accordingly.

A smishing detector that uses a distilBERT base uncased model. I trained and deployed the model with Amazon SageMaker and served it via a lambda function. The experiments and complete project report are available on the Git repository.