Introducing the 2019–2020 Oasis Fellows!
We’re excited to announce our 2019–2020 class of Oasis Fellows. The Oasis Fellowship is a program designed to support outstanding students doing research and development in distributed systems, security, privacy, and machine learning.
As a part of the program Fellows receive funding and support for their research initiative, an opportunity to work with the Oasis Labs team, and an internship here in the Oasis offices in San Francisco.
We’re so excited to have these incredible Fellows join us soon and can’t wait to see their research in action.
More about each Fellow below.
Vasilis Mavroudis
Tell us a little bit about yourself?
I am a Doctoral Researcher in the Information Security Group at University College London, advised by Prof. George Danezis and Prof. Emiliano De Cristofaro. My research focuses on Machine Learning, Blockchains, and Privacy, while I have also pursued projects on Trusted Hardware, Privacy-preserving Computations, and Technical Market Manipulation. Currently, I am investigating deep neural networks and their security/privacy implications. In the past, I have worked on problems in (permissionless) blockchain consensus protocols, supply-chain threats against integrated circuits and side-channel attacks in anonymity networks.
For more information please visit my website: https://mavroud.is
What are you working on as a part of the Oasis Fellowship?
Privacy-enhancing technologies (PETs) are frequently used to mitigate privacy threats such as protecting personally identifiable information and communication anonymity. However, such technologies can also enable novel computational scenarios and applications that would otherwise be impossible without clear privacy guarantees. As part of the fellowship, I will work on PETs that allow mutually distrusting parties to collectively compute functions while disclosing only the minimum amount of information in the process. Overall, my goal is to improve the trade-off between the benefits of collaboration and the risks associated with sharing sensitive information.
What made you excited about working with Oasis?
The Oasis blockchain-based platform will provide an ideal basis for my research and will enable me to explore new computing paradigms based on decentralized trust. Moreover, I am very excited at the prospect of collaborating with Oasis’ researchers to design new schemes, advance the capabilities of the platform, and enable new cloud-based, real-world applications.
Savvas Savvides
Tell us a little bit about yourself?
Savvas Savvides is a Ph.D. candidate in the Department of Computer Science at Purdue University under the supervision of Prof. Patrick Eugster. He received a B.Sc. in Computer Science from the University of Manchester, UK, and an M.S. in Computer Science from New York University, USA. His research interests span the areas of information security, distributed systems, and cloud computing with an emphasis on secure and efficient distributed computations. As part of his research at Purdue, Savvas has a) proposed a novel data type abstraction which enables the use of Partially Homomorphic Encryption (PHE) in ways that improve performance and security, b) introduced two novel symmetric PHE schemes that are more efficient than state-of-the-art asymmetric PHE schemes without compromising on homomorphic expressiveness and c) proposed a programming model which leverages multiple security mechanisms such as PHE, trusted execution environments and client-side computation to achieve efficient secure computations in the cloud. Savvas has also completed an internship at IBM T. J. Watson Research Center, as a Research Summer Intern where he worked on estimating the execution-time of Spark applications through static program analysis and runtime monitoring and another internship at Fortanix as a Security Research Engineer Intern where he worked on achieving secure consensus using Intel SGX.
What are you working on as a part of the Oasis Fellowship?
As part of my fellowship with Oasis Labs, I will be working on adding support for secure multiparty computation and secure consensus on Keystone.
What made you excited about working with Oasis?
Oasis Labs tackles the issue of preserving user privacy in a practical manner which is, in my opinion, one of the biggest challenges facing the field of computer science in the 21 century. Preserving the confidentiality and integrity of computations as well as making services more available and more performant are notions central to my research interests. For this reason, I have chosen to work for Oasis Labs and I am looking forward to collaborating with other like-minded scientists who recognize the importance of security.
David Dao
Tell us a little bit about yourself?
I’m a Ph.D. candidate at ETH Zurich DS3Lab and researcher at Stanford University working on safe AI for Sustainable Development. My research leverages Blockchain-based incentives and AI to realize new types of privacy-preserving learning systems for the environment, human well-being, and ethical use. I’m the founder of GainForest, an award-winning non-profit that provides AI-powered conservation tools to prevent deforestation. I was an engineer in Silicon Valley and a research fellow at Berkeley AI Research (BAIR) and Broad Institute of MIT and Harvard. A Global Shaper at World Economic Forum, I organized several large conferences in Germany, Silicon Valley, and at Harvard. My work was featured in MIT Technology Review , The Scientist, The New York Times and at United Nation’s Climate Change conference.
What are you working on as a part of the Oasis Fellowship?
I work on Kara, a privacy-preserving data marketplace for health data that aims to improve medical knowledge by unlocking the vast amount of silo-ed medical data for research.
In cooperation with doctors from Stanford Medical School, we study the computational economics and incentives around private patient data and develop private AI algorithms and systems that have the potential to save lives in the future without spilling medical secrets.
What made you excited about working with Oasis?
Privacy is critical for Kara. Oasis not only provides the strongest privacy model I’ve seen so far in practice, but also allows for high-performance computation on a decentralized ledger that enables us to train AI and study our economic questions (all while keeping the patient’s data secret). In fact, its strong security guarantees are a big reason why a recent clinical trial proposal at Stanford got approved by regulators. We are now looking forward to collect real patient data and work together with the team at Oasis.
For more information about Oasis go to www.oasislabs.com or email us at info@oasislabs.com.
