Cryptographer · Engineer · Builder
Building open-source tools that prove things are real.
I’m an 18-year-old developer from Brazil, CTO at a GovTech company serving municipalities across the country, and creator of open-source cryptographic tools. I’m writing a paper for IEEE WIFS 2026 on sensor-based screen recapture detection , a method with no published precedent. This fall, I’m moving to Granada, Spain to study Computer Engineering.
Yorubs
Gamified mobile app to learn the Yorubá language. 11 learning modules, 19 exercise types, proprietary AI tutor, Sacred Library of 19 Orixás, and 358 authentic proverbs. Built with React Native and Expo.
Amrntech
GovTech platform serving municipalities and state governments across Brazil. Digital systems for health, education, security, infrastructure, social assistance, environment, civil defense, and transport.
Founded Amrntech
Started building GovTech solutions for Brazilian municipalities. Digital transformation for public services.
Created Yorubs
Built a gamified Yoruba language learning app with proprietary AI tutor and Sacred Library.
Became CTO at Amrntech
Leading engineering across health, education, security, and infrastructure systems for state governments.
Launched TrueShot
Open-source cryptographic photo verification. ECDSA P-256 signing, 14-sensor fingerprinting, screen recapture detection.
IEEE WIFS 2026 Paper
Writing a paper on sensor-based screen recapture detection. First non-visual approach with no published precedent.
Moving to Granada, Spain
Starting Computer Engineering at the University of Granada. New chapter.
Sensor-Based Screen Recapture Detection
A non-visual approach to detecting photos of screens. Cross-correlates 10 physical sensor signals at capture time. Independent of image content and display technology. First method that scales with display improvements rather than degrading.
Post-Hoc Cross-Device Corroboration
Verify that multiple photos from different devices depict the same real event , without any communication between devices. Uses barometric pressure, timestamps, GPS, and sensor coherence. Attack cost scales exponentially with each additional device.
Screen Recapture Detection Results
Redmi Note 11S · Purely sensor-based · No ML · Threshold: 60
Mobile
Backend
Frontend
Cryptography
DevOps & Tools
Open for collaborations, research opportunities, and interesting conversations.
© 2026 Yuri Borges Martins · Built with Next.js, GSAP & Three.js