Making AI agents secure and auditable.

Hi, I'm Uilton. I'm an AI agent security & governance researcher and technical leader with 25+ years building products and technology companies. My work asks what security guarantees remain valid when AI agents are trusted with confidential documents, delegated credentials, and untrusted content in the same session. Currently CTO at Prodeal and a doctoral researcher at Universidade Feevale.

Recent talks

Abstract Presentation

Conversion of PDFs into Structured Data for Auditing Projects under the Paulo Gustavo Law

Inovamundi 2025

Presented a methodology for extracting and structuring data from PDF documents to support auditing processes for cultural projects funded under Brazil's Paulo Gustavo Law.

Inovamundi 2025 Workshop Photo
Workshop

AI Agents with Python and Generative AI

BAR 2025

Hands-on workshop covering the development of AI agents using Python and generative AI technologies.

BAR 2025 Workshop Photo

What people say about me