Covalent, the privacy-preserving data utilization protocol: a project overview

Full disclosure: This article is not intended as investment advice. It is just my personal opinion about the Covalent project. You should always do your own research. I am part of “the article group” which rewards me for writing this article and supports me for ventilating my own personal opinion.

Covalent is a privacy-protected, decentralized computing platform and protocol. Covalent’s aim is to provide a distributed network computing layer that is able to extract/analyse information from data without compromising privacy. Within the Covalent ecosystem the COVA token will be used to compensate data owners for each micro usage of their data. This may sound complicated, but let’s use the following example to illustrate a typical use-case of the Covalent protocol. Let’s assume that you have your own medical data stored in a secure place and you are reluctant (and rightly so) to share this data with third parties (e.g. research institutions) due to privacy concerns. Using the Covalent protocol relevant information can be distracted from your medical data without compromising privacy and without any third-party ever gaining direct access to the data itself. The extracted data can then be aggregated and analysed, hopefully leading to new insights in how to prevent/treat certain diseases. As a reward for sharing your data, you will be compensated in COVA-tokens and have helped advance medical research at the same time.

In 2017 a total of 22 zettabyte (ZB) of data was created (1 ZB = 1,000,000,000,000 gigabytes). However, out of this huge amount of created data only 1% was actually being utilized. Data remains unutilized since sharing the data often results in compromised privacy. Once data is shared the data owners expose their sensitive data and lose all control of its usage and distribution. So, in that regard it is understandable that only a fraction of the potentially available data is actually being utilized.

What we currently see in this space is that areas where data is aggregated naturally (e.g. online search and online advertising) machine learning and AI have progressed in a rapid pace. Just look at how Google and Facebook make use of aggregated user data and algorithms to determine optimal search and advertisement strategies. Yet in other areas (e.g. healthcare and education) the progress is very slow. One of the reasons why progress is sluggish in these areas is that the data is more sensitive and fragmented and as a result there is often less data available to feed the AI models and as a result they are limited in their effectiveness.

From a machine learning and AI perspective it would be ideal if data could be shared so it can be aggregated into large data sets (with more data often leading to more reliable models), which can be accessed by various authorized third parties; instead of being fragmented and siloed. However, two major problems arise in this scenario. 1) Privacy, data owners are reluctant to trust third parties with handling their data (especially when it consists of sensitive information, such as medical records). 2) Value network, data owners are not incentivized to share their data. This is where blockchain technology and especially Covalent comes in. By leveraging blockchain technology, Covalent allows for private modeling and computing without compromising the privacy of the data providers and through their COVA tokens compensate data owners for each micro usage of their data.

In order to build a privacy-protected, decentralized computing platform and protocol, Covalent will implement the following features: Covalent Virtual Machine (CVM), Trusted Execution Environment (TEE), and a Data Marketplace. In this section I will briefly touch upon these features, as it would be too lengthy to provide an in-depth analysis of each technology. In addition, Covalent has published its own deep-dive article explaining Covalent’s technology (I found it to be a very comprehensive and well-written article).

Covalent Virtual Machine (CVM): Covalent is building an enhanced version of the Ethereum Virtual Machine, which will be capable of large scale-computations. Within the CVM, a computation is preformed off-chain by a single node through a smart contract, with all the other nodes only needing to verify a proof of the computation. Since the process of verifying a computation is inherently faster than repeatedly performing the computation, the speed of computing is largely increased; hence making large-scale computation possible. In addition, Covalent has developed Centrifuge, a language for writing Smart Policies that is designed to take advantage of the enhanced functionality of the CVM.

Trusted Execution Environment (TEE): To facilitate privacy of computation Covalent will make use of a TEE, which can be described as an isolated environment that runs in parallel with the operating system and provides security for the rich environment. It guarantees that data and code that is loaded inside the environment to be protected with respect to confidentiality and integrity. A TEE, as an isolated execution environment, provides security features such as: isolated execution, integrity of applications executing with the TEE, along with confidentiality of its assets.

Data Marketplace: Covalent will also be integrating a data marketplace where data owners, model trainers and SGX host miners can interact with each other. Data owners can list their private datasets on a pay-per-use basis. Model trainers pay for the use of datasets that are available on the marketplace to run their private models. SGX host miners facilitate the transactions (and be compensated with COVA tokens) while satisfying the privacy and proof of computation requirements.

In short, the CVM relies on a TEE, to make large-scale computation compatible with smart contracts. While the SGX Host Miners perform typical functions like transaction verification and block production, they also host the secure environment for off-chain computation. Each computation is performed off-chain on a single node to maximize privacy and reduce on-chain network load. When a computation task is completed, the off-chain node submits proof of computation together with the reported output results to the smart contract protocol so that the other nodes in the network can verify the computation.

Vincent Li (Co-founder) is a former doctoral candidate at Harvard University, but decided to pursue a career in finance instead of obtaining its PhD. Vincent has worked as an engineer for Tower Research Capital for 1 year, then moved on to Citadel, which is a leading hedge fund. After working at Citadel for 2 years he obtained a position at a start-up for freelance software developers called Gigster, where he worked for 1 year.

Raymond Gao (Co-founder) obtained a Master’s degree in mechanical engineering from Princeton University, where he also spent two years as a Research Assistant. In his bio it states that he is an early BTC and ETH adopter and miner. In 2016, he joined FreeS Fund where Raymond led and participated in the investment of big data and Internet of Things companies.

Shundan Xiao (Software engineer) obtained a Master’s degree in computer science from North Carolina State University. Shundan has over 5 years of experience as a veteran coder and software engineer at companies such as: Amazon, LinkedIn, Gusto and Gigster.

According to their LinkedIn page, the Covalent team consists of at least 9 other members, hence for more information about the other team members you can check their LinkedIn profiles as well.

On their website Covalent mentions two advisors, namely: Shuoji Zhou (founder of FBG capital) and Jia Tian (Investor of Bitfinex and limited partner of Bitfund). In addition, Covalent has already received investments from several well-known Venture Capitalist totaling over 10,000,000 USD. Some of these VC’s include: ZhenFund, FBGCapital, IOST/ Bluehill and Node Capital. Covalent also received funding from Huobi and San Francisco based cryptocurrency fund Long Crypto Group (Source).

Currently the Covalent project is still in pre-ICO stage and dates of the ICO have not yet been disclosed. The token to power the Covalent eco-system will be called COVA and it will initially be launched as an ERC-20 token based on the Ethereum blockchain. When Covalent’s mainnet is released (no date set yet) the ERC-20 COVA tokens will be swapped for native COVA blockchain tokens.

If you wish to be notified when more information about Covalent’s ICO becomes available I would advice you to subscribe to Covalent’s Twitter and join their Telegram channel.

The covalent project is still in an early stage and some information, such as ICO details, are still lacking. Also, Covalent’s whitepaper is pretty daunting to read and more targeted to people with prior knowledge in the realm of machine learning and blockchain technology, rather than to the “average” person who lacks this knowledge. That being said, I think that the Covalent team is doing a good job with their deep-dive articles trying to explain the technology behind the Covalent project in more layman terms. I think that the idea to develop a privacy-protected, decentralized computing platform is promising and this is definitely a project that I will add to my short list of projects to follow during the coming months and see how this project and the ICO details will unfold.

Sources for more information on Covalent:

Website: http://covalent.ai/index.html

Whitepaper: https://docsend.com/view/dvvb75n

Telegram: https://t.me/covalentofficial

Twitter: https://twitter.com/covatoken

Blog: https://medium.com/@covatoken

Github: https://github.com/covalent-hq

Subscribe to my channels: Medium, Twitter and Steemit if you enjoy my articles and would like to be informed about blockchain, cryptocurrency projects and news. You can also read my articles on LinkedIn.  If you have any questions about this article, please comment in the comment section below. Claps are also welcome, Thank you!  CryptoShowdown