Unlock to reveal name

Machine Learning Engineer

  • Profile Created:
  • 3 years ago




  • Current Company 2019 - Present

    Computer Vision / Machine Learning Engineer

    My job was to improve ID Card Validation and OCR process, using machine learning. Being the only machine learning developer, I took the following projects from zero to production: Developed a deep learning based module to detect and rotate a card in a given image Developed a deep learning based module to distinguish between a photo of a portuguese ID and a photo of a printed portuguese ID, as well as from a photo of a photo of a portuguese ID (photo from computer screen for example). Developed a deep learning based module to do card segmentation of portuguese ID cards. Developed a deep learning based module to detect ID card security elements, for fraud check. Previously I was continuing HealthyRoad project (Machine Learning and Computer Vision in Driver Monitoring System).

  • HealthyRoad 2013 - 2019

    CTO and Co-Founder

    HealthyRoad is a company that develops innovative products that allow to solve real problems through the most advanced biometric analysis technologies using machine learning and computer vision based on a camera. As Chief Technology Officer, I am responsible and main developer of all the development of the projects created in HealthyRoad. HealthyDrive: Development of an application to detect and prevent drowsiness and fatigue in a person with real-time video processing. C++ application using OpenCV in embedded Linux (Odroid and NVIDIA Jetson), that also communicated with a cloud platform in order to send data and alerts. Fleet managers could later check their drivers data. HealthySecurity: Development of an application for monitoring security guards. C++ Desktop application using Qt and OpenCV in Windows that: Recognizes the security guard (to unlock the computer); Detects if he leaves the station (locking the PC); Detects if he is inactive for a defined period of time; Detects if he fall asleep; Create reports with the work statistics of the security guards. HealthyRoad SDK: Creation of a library that wraps all the biometric functionalities developed at HealthyRoad: Drowsiness and Fatigue Detection; Stress Quantification; Emotions Classification; Facial Identification; Vital Signs Measurement (HeartBeat). Main technologies used in HealthyRoad: C++ OpenCV Machine Learning www.healthyroad.pt Porto, Portugal

  • Current Company 2018 - Present

    Full Stack Engineer and Co-Founder

    By containing a series of exercises, children can train their orofacial motricity anywhere in a fun way! It also keeps a record on when they exercise, and how good is their evolution. Technologies used: react-native (javascript) realm

  • AskForEat 2019 - 2020

    Full Stack Engineer and Co-Founder

    AskForEat is a mobile app to easily order at a bar or restaurant. It allows the customer to order and pay on the table through the app. By scanning a QRCode on the table, the restaurant knows where the order came from, and if the order is already payed or not. Technologies used: App: react-native; Backend and Frontend for the restaurant dashboard: nodejs AngularJS mongodb

  • FreedomGrow 2011 - 2012


    Development of a web application for monitoring wireless sensor networks. Technologies used: C# ASP.NET Microsoft SqlServer


JavaScript (react-native, react, nodejs backend)