Courses: Deep Learning, Natural Language Processing, Computer Vision, Reinforcement Learning, Computational Statistics, Formal Software Specification, Software Architecture, Data Science, Information Visualization.

Thesis: Towards Reliable Counterfactual Fairness Audits in Real-World Machine Learning Applications. Advised by Prof. Chrysoula Zerva and Dr. André Carreiro.

Thesis Grade: 20/20 · Overall Grade: 18/20

Selected courses: Systems and Signals, Statistical Signal Processing, Audio and Image Processing, Numerical Methods and Applications, Design and Implementation of Intelligent Electronic Systems.

• Co-founded a logistics tech startup focused on optimizing middle-mile supply chain operations through predictive modeling and vehicle allocation.
• Led the technical roadmap from ideation to deployment, delivering a high-stakes predictive solution for a major Brazilian carrier within a 3-month deadline.
• Bridged engineering and business value by communicating technical solutions to non-technical stakeholders.

• Developed an extension to Counterfactual Fairness that relaxes the causal graph requirement and measures uncertainty due to causal assumptions. Presented at ECML PKDD 2025 Bias Workshop.
• Proposed desiderata for counterfactuals for Counterfactual Fairness inspired by synthetic data generation literature.
• Introduced benchmark strategies for Causal Discovery algorithms with more realistic mechanisms and domain knowledge constraints.
• Worked with causal diffusion models and causal autoregressive normalizing flows.

Developed business insights and automated processes using Python and SQL for invoicing, billing, reconciliation, safeguarding, deferrals, and FP&A in the acquiring sector.

Produced business insight using Python in the Order to Cash chain in the food sector, to reduce transportation costs.

• Developed control systems to analyze the frequency response of transducers in lung ventilators.
• Co-developed a neonatal lung simulator with variable compliance.

Led a 32-member team for the development of 49 projects of web and mobile applications. Semester revenue of BRL 730k (EUR 117k).

D. Valério, C. Zerva, M. Pinto, R. Santos, A. Carreiro. Counterfactual Fairness with Graph Uncertainty. Proceedings of the 5th Workshop on Bias and Fairness in AI at ECML PKDD, 2025.

S. Mello, A. Lino-Alvarado, D. Valério, C. Estevam, M. Dias, K. Barros, A. G. Ferreira Junior, H. T. Moriya. Development of a Neonatal Lung Simulator with Variable Compliance. IX Latin American Congress on Biomedical Engineering and XXVIII Brazilian Congress on Biomedical Engineering, 2024.

S. Mello, D. Valério, J. Angelo, R. Romano, H. Moriya. Development and characterization of a system for measuring and analyzing the frequency response of transducers used in the ventilation of neonates and premature infants. 12th Brazilian Metrology Congress, 2023.

Co-developed Voice-Over Vision using LLMs and video processing to create real-time, contextualized audio descriptions for YouTube videos. Designed to improve web accessibility for visually impaired users. Won 2 hackathons.

Workshop on Fairness in AI by the Center for Responsible AI (CRAI).