04/11/2022 - Talk by V. Papakonstantinou & A. Angelidakis, Mobiltron "Authenticating individuals using behavioral biometrics and machine learning"

04/11/2022, 18:30 Athens time, remotely via Zoom and on Campus 145.Π58


04/11/2022, 18:30 Athens time
on Campus 145Π58

via Zoom:
Meeting ID: 953 8417 5459
Password: 198509


Authenticating individuals using behavioral biometrics and machine learning


Inspired by a friend's fatal accident, we set out to develop a system to predict potential emergencies and help save countless others. The core idea of the system was to build a behavioral baseline for each individual and monitor for signals that could imply potential emergencies in a given context. While creating the original user baseline, we realized the value of having verified user-generated behavioral data. The traditional process of active authentication, based on conventional credentials or physiological biometrics, which we continue using every day, was insufficient for this case. Thinking on first principles, we developed a novel system to solve our problem, which eventually became a platform that can passively authenticate individuals through their mobile devices using behavioral biometrics data and machine learning.

About the speakers

Vassilis Papakonstantinou
Vassilis is a technology entrepreneur with over 25 years of hands-on experience across various industry sectors. He likes building things, exploring ideas, and connecting people. He is a co-founder and partner of Blue Dome Capital, an investment manager seeking to capture emerging opportunities in technology transfer and innovation-driven ventures in Central & South-Eastern Europe. He is the co-founder of Mobiltron, Inc, a technology company developing a next-generation user authentication platform using behavioral biometrics for mobile applications. He is a partner of Sylipsis, Inc, a company helping researchers move their ideas from the lab to the market and providing managed technology development as a service. Vassilis consults the Boards of various organizations on issues related to technology and innovation. He currently sits on the Board of Fasmatech, a startup developing mass spectroscopy technology, and VIORYL, a chemical engineering company. Previously, he was on the Board of Metis Cybertechnologies, a company developing vessel performance solutions, and helped the investors achieve a 3X exit. He is the co-founder and Vice-Chairman of the MIT Enterprise Forum Greece, the Greek chapter of an MIT-inspired global network, enabling technology entrepreneurs to rapidly transform ideas into world-changing companies. He is the co-founder and Treasurer of the Hellenic Innovation Network, Inc, a Boston-based charity that supports the entrepreneurial growth of the Greek and Cypriot tech ecosystem. He holds degrees in Mechanical Engineering from NTUA and Ocean Systems Management from MIT. He has conducted research in fluid mechanics, control systems, risk management, and system dynamics.

Angelos Angelidakis
Angelos is a Data Science professional with +7 years of experience in descriptive and predictive analytics, machine learning, and artificial intelligence. He is currently working at Accenture as a Data Scientist. Before that, for three years, he was with EY, a CESA Data & Analytics Center of Excellence member, in innovative solutions based on Data Science and Data Engineering. He started his career as a Data Scientist for Mobiltron and eventually became its Chief Data Scientist, developing significant parts of its technology. Earlier, he has a research assistant at the Telecommunication Systems Research Institute, dealing with reinforcement learning. He holds an M.Sc. in Artificial Intelligence from the Technical University of Crete, as well as an M.Eng. in Computer Science and a B.Eng. in Electronic and Computer Engineering. Angelos has worked in various industries, including Telco's, Supply Chain, Car Fuel, Pharmaceutical, Public Sector, Food Service, and Facilities Management, and has functional expertise in Artificial Intelligence, Machine Learning, Predictive Modelling, and ML Ops.