Dbrain Rating Review
We assign the Dbrain project a "Stable" rating.
Dbrain is a collective platform for creating applications based on artificial intelligence. The platform connects ordinary users and data scientists to help create real AI solutions for business based on raw data, simplifying this process for all participants. Dbrain enables anyone who has a mobile device or a computer to perform basic tasks for labeling up data and receive the project’s internal DBR cryptocurrency in exchange for this. Dbrain also enables data scientists to participate in development contests. The process creates a ready solution available for business after training neural networks on this data.
Dbrain is staging its ICO with a finished alpha version of the platform, including a web application and Telegram bot. Users are solving the labeling task in test mode, give feedback on the service, helping develop the platform and getting paid for this.
The Dbrain team consists of industry-renowned professionals who have several successfully launched AI-based products between them, including Connectome.ai, R-SEPT and the Icon8 chatbot. The advisory board also consists of experienced professionals in the fields of AI, venture capital investment and the crypto industry.
We highlight the team’s competence and the promise of the product being created; we see its relevance in a fast-growing market. At the same time, we must also point out a lack of clarity and transparency of information regarding the token economy and distribution of the tokens. We also see weaknesses in the marketing strategy. A large share of the total issue of tokens is allocated to the team and the reserve fund. The token price will be affected by a large number of factors, many of which are negative in the medium term.
We should also mention the risks inherent in the platform, including the complexity of the product, the legal risks, as well as the large number of competitors.
General Information about the Project and ICO
Dbrain is an open blockchain platform that links crowdworkers and data scientists to help businesses sort and label data, and create an AI solution based on this. A crowdworker is any person who performs simple labeling and checking tasks, for which he will be rewarded with the project’s internal cryptocurrency in the future. Data scientists create solutions architecture and provide training based on labeled data. Various commercial or state companies can use the AI solutions or data obtained, for the optimization of business processes, product improvement and other tasks.
Dbrain implements a convenient labeling and data processing mechanism for interacting with crowdworkers via a proprietary web application, mobile application and Telegram bot, making it easy for anyone who wants to become a member of the project and receive rewards in DBR tokens.
The platform is designed to link the fast-growing demands of businesses for manual AI training with the huge labor resources available worldwide. The ability to generate income from cryptocurrency in exchange for performing simple labeling and data processing tasks will be a strong point for supply worldwide, especially in countries with small wages and high unemployment. The synergy of these two parties will enable AI to attain a new level of use by various companies.
The platform is helped by the fact that it has an MVP; it already gives tasks to the community and receives feedback for them. The project is staging an ICO to create an ecosystem with a proprietary token and to finish the product and complete its marketing.
The project sold tokens at the Seed stage from 01/01/18 to 01/02/18, where it raised $2,500,000.
ICO Private Sale start date: 15/04/2018
ICO Private Sale end date: 01/07/2018
Token name: DBR
Hard cap: 17,500,000 USD
ICO price: 1 DBR = 1.6 - 2 USD
Minimum Buying Transaction: $50,000
Maximum Buying Transaction: $2,000,000
Unsold tokens will be burned.
Size of emission: 40,000,000 DBR:
40% - Unrestricted sales to investors at the ICO
30% - Team
20% - Reserve fund
10% - Community
The funds raised will be spent as follows:
55% - Product marketing
35% - Product development
7% - Community management
3% - Legal
During the ICO, the project is offering bonuses to participants depending on the amount invested and the status of the relationship. For example, with an investment of $50,000 an investor can count on a bonus of up to 10%, 15% for $200,000 and strategic partners and institutional investors can obtain up to a 20% bonus.
The project has a flexible system for lock-up periods; partial unlocking for all participants is provided according to the schedule: 25% will be available immediately after the ICO, 25% of the amount will be unlocked every 3 months and the final package will be fully available after 9 months.
The team’s tokens have a lock-up period of 12-15 months.
Description of the Project Services
Dbrain is an open platform for creating ready-made solutions based on artificial intelligence for business, which brings together ordinary users as a learning force and data science developers for the purpose of creating AI-models. Users are rewarded for their work with DBR tokens.
The platform offers benefits to all its ecosystem participants. Crowdworkers will have an opportunity to earn money for the verification of images and the development of AI. Businesses will have a wide range of integration and customization of solutions with individual parameters.
In general, the process for using the service is as follows:
Data owners download and place data sets on the platform.
Crowdworkers label the data and receive DBR tokens in exchange for their work.
Data scientists train neural networks and create AI applications, receiving tokens as a reward.
Businesses receive the opportunity to use existing applications in exchange for platform tokens or order new models for their needs.
Let us consider the services for each category of participant in more detail:
Ordinary users can become crowdworkers and be paid in DBR tokens for manually labeling data, i.e. determine what is shown in the picture.
Users can mark and label data using a Telegram bot, a web interface and a mobile application (in the near future).
The Telegram bot enables a large number of users, even those without access to banking services, to receive tokens in exchange for image processing. Tasks are usually simple and their quality of performance affects future orders, as well as the number of remuneration tokens.
The team already has extensive experience in creating Telegram bots. Previously they developed the successful Icon8 chatbot that reached third in Facebook rankings.
The web application and mobile application integrated with Ethereum DApps enables solving more complex tasks involving labeling data with an intuitive interface. A personal account provides token accounting and completed tasks for users.
The platform has a working MVP, where users can perform image processing and receive DBR tokens for this.
For companies and data science developers
Companies can pay for crowdworkers’ and data scientists’ services with tokens, and receive a trained neural network and a working AI application to solve their business problems in return.
The Dbrain platform is designed to provide a scalable and accessible infrastructure in order to give enterprises access to services with a high level of AI development, conveniently integrated through the API. Dbrain will offer services, integration within businesses and customization possibilities.
For AI developers, barriers to creating commercial products using AI are significantly reduced, and access to accumulated data sets, unique data providers, business partners and a distributed pool of crowdworkers that label existing data in real time are provided.
The platform’s proposed services look quite promising; they are thought out and are already partially functioning, giving the project an additional advantage at the ICO. We stand with the project’s assessment of the prospects for AI solutions, and we also positively assess the development of the potential market for the Dbrain project.
The artificial intelligence market is growing at a fast pace. Many companies are planning to use AI technologies in their business in one way or another. This is due both to a high evaluation of this technology from experts, and a trend for the use of AI in various fields. According to Bloomberg, current mention of "AI" on quarterly earning calls overtook that for "Big Data".
By the end of 2018, AI technology will bring $1.2 trillion of added value to the global economy, and AI solutions for customer-related operations will be one of the most important tasks for business.
According to forecasts, the value of business in the AI sphere will reach $3.9 trillion by 2022.
According to PwC, artificial intelligence could add $15.7 trillion to global GDP by 2030 as a result of the introduction of this technology to business.
PwC has also ranked regions in which the contribution of AI technology will be the most significant:
"AI promises to be the most disruptive class of technologies during the next 10 years due to advances in computational power, volume, velocity and variety of data, as well as advances in deep neural networks (DNNs)," said John-David Lovelock, research vice-president of Gartner. "One of the biggest aggregate sources for AI-enhanced products and services acquired by enterprises between 2017 and 2022 will be niche solutions that address one need very well."
These needs can include methods for improving customer relationships, attracting new revenue streams and means to reduce costs, either operating costs or maintenance costs for existing products.
Companies including Google, Apple, Microsoft, IBM and Nvidia are already actively involved in the research and development of AI-based products and services. According to CB Insights, start-ups are beginning to specialize in artificial intelligence with a focus on industries such as customer relationship management, automotive, sales, marketing and commerce.
One application of thia technology is the use of virtual assistants which can take on simple customer requests and tasks from call centers, reducing costs of customer service. Ordinary human operators can freely devote their time to more complex issues which in turn can also improve customer service.
The use of deep neural networks will enable organizations to perform intelligent data analysis and pattern recognition in huge data sets that cannot be recognized and classified any other way. Such opportunities have a huge impact on the ability of organizations to automate decision-making and interaction processes.
The most relevant and/or promising industries where AI is expected to be extensively used by 2020 are as follows:
Media and advertising
Automotive industry and transport
Oil and gas
The total market was estimated at $12.5 billion in 2017; it will reach more than $46 billion in 2020 while maintaining growth forecasts.
We evaluate this market as dynamic and promising. The DBrain project needs to occupy 0.5-1% of the global market for a potential revenue of $63-125 million per year following general market dynamics. We consider this scenario to be the base for the project.
The market is rather competitive.
According to Venturescanner, there are 2158 companies in 13 categories in 71 countries, developing in the field of AI.
In addition to the fact that there are many companies dealing with AI for various purposes, there are direct competitors that offer users data labeling in exchange for rewards. Let us consider some of the most important competitors in this market:
Amazon Mechanical Turk - one of the strongest competitors in the market. It is a platform for coordinating the use of human intelligence to perform tasks that computers cannot perform. This is one of Amazon Web Services’ sites. Employers can post assignments, such as choosing the best among several store photos, describing products or identifying artists on music CDs. Employees can view potential vacancies and fill them in exchange for a reward determined by the employer. Requesting programs use an open application programming interface and the MTURK Requester for placing jobs.
Yandex.Tool – this platform provides the opportunity to receive remuneration online, performing simple tasks that computers cannot yet cope with. Basically, users analyze and evaluate different content, for example, checking the relevance of sites to search queries, comparing pictures and determining categories of products. A prospective employee requires a computer or a mobile device with internet access and some free time.
Figure Eight — an AI company based in San Francisco, that attracted $58 million in venture capital. Figure Eight uses human intelligence to perform simple tasks such as text decoding or recognizing images to teach machine learning algorithms. Figure Eight automates tasks for machine learning algorithms that can be used to improve search results in catalogs, photo recognition or customer support. The technology can be used to develop autonomous vehicles, intelligent personal assistants and other technologies that use machine learning. The company is working with companies such as Autodesk, Google, Facebook, Twitter, Cisco Systems, GitHub, Mozilla, eBay, Toyota and others.
The market is rather competitive, the above companies are not the only ones present, but they have the strongest impact and they are similar in functionality to Dbrain. The Dbrain platform differs from competitors in that data science developers can also become participants on the platform.
We see a highly competitive environment and some similarities in approach. Successful competition with large, established companies is possible only with the creation of a high-level product and an effective marketing strategy.
Let us consider the main members of the Dbrain project team, their positions and skills.
Chief Executive Officer
Dmitry has extensive experience in managemen, as well as data development. He was the founder of Icon8 chatbot and Flocktory, a start-up sold to QIWI for $20 million in 2017. Has more than 13 years of experience related to data or management in one way or another.
He graduated from the Moscow Institute of Physics and Technology, receiving an additional financial education at the Russian Economic School.
Chief Technical Officer
Aleksey is the founder of connectome.ai and R-SEPT. The connectome.ai project carries out search and implementation of business solutions in the field of AI. Aleksey has more than 3 years management experience in the field of R & D for IT and Artificial Intelligence.
Chief Design Officer
Founder and CEO of Wikipedia of Things, Thngs, graduate of the Strelka Institute for Architecture, Media and Design. Dima has more than 12 years’ experience in product design and development.
From our point of view, the project management team is sufficiently competent in its field. Its members have experience enabling the creation of a product for the artificial intelligence market based on the strengths and weaknesses of solutions currently available. The team is also able to offer its own solutions to the market. The competitiveness of the product can only be assessed together with its marketing and technical component; however, we consider its potential to be adequate.
The team includes other specialists who have posts that presuppose management, programming and other functions. However, it is not possible to assess their experience, nor that of some members of the core team.
The project’s advisors are experienced and to some extent well-known in the market, and they offer a variety of relevant skills. The participation of high-level professionals demonstrates the seriousness of the project and the esteem in which it is held by the industry.
Key advisors to the project include:
Yobie Benjamin - CTO of Token.io, has extensive experience; influences regulation processes and the development of payment systems. He is a member of the Federal Reserve Bank’s expedited group on accelerated payments. A venture investor and great expert in the field of fintech.
Eleanor 'Nell' Watson - engineer, educator and philosopher. Nell was the initial founder of Poikos (now QuantaCorp). The company's proprietary patented technology enables fast and simple 3D body measurement from only two planes (front and side), using a 2D camera, by sophisticated deep machine learning techniques. This service enables quick and accurate service personalization for telemedicine, retail, etc. Nell lectures in the field of artificial intelligence, the philosophy of machine learning, human-machine relationships and the future of human society worldwide.
Vadim Fedchin - Master of Law and Finance at Oxford University. Member of the Open Data Institute (UK), has extensive experience of venture investments in the field of artificial intelligence.
Ed Gurinovich - Founder of Carprice, Carmoney and MyTime. Venture investor in artificial intelligence.
Dean Patrick - Founder and Managing Partner of G2H2 Capital and a graduate of Stanford University. He is an experienced investor in digital technologies and cryptocurrencies.
Mikhail Larionov - CTO of BitGuild, has work experience with Facebook (Messenger Platform) as well as for Disney Interactive.
Other equally experienced and important experts in the fields of finance, cryptocurrency and data management, details of whom can be found on the project website.
The project involves a large number of advisors, which could increase the likelihood of implementing a sophisticated product drawing on their experience and assuming close interaction with the team.
The project declares many partnerships, including with eminent companies such as Microsoft and Nvidia, but it is not possible to assess these partnership relations and their benefits since details are not disclosed.
The platform has developed its own DBR token to facilitate interaction between participants and creating a remuneration system for crowdworkers.
DBR tokens will be units of account within the platform. Commercial companies will need tokens to create an order on the platform and pay for it. Data providers and crowdworkers are able to earn tokens for their work.
The interaction of the two sides - demand from business and supply from crowdworkers and data providers - can create a balanced interaction model, if the token can be easily purchased on exchanges or the platform itself, i.e. with proper liquidity.
The emission and nature of the token is down to the need for blockchain technology in the framework of interaction between participants. The project uses 2 proprietary technical solutions:
SPOCK – this creates a consensus algorithm among users when labeling, and checks the quality of labeling.
PICARD – this provides data security and automatic selection of the best model, as well as payments for the use of API solutions in tokens.
Thus, a fairly autonomous system for order distribution and future revenues is created, featuring confidentiality and data security.
Dbrain receives commission revenues from each transaction within the platform: Data labeling, the development of AI applications, and API usage for AI applications hosted on the platform. The commission level is 10%. Payments for data labeling and development of AI applications are performed immediately.
Only 40% of the total issue is allocated to the coin sale during the ICO, whilst the team and the community will take 40% and the remaining 20% will be allocated to a reserve fund.
We think that the current distribution model could be perceived ambiguously by investors, despite the long lock-up periods involved. In particular, we cannot see how the reserve fund that will benefit from half of the proceeds of the coin sale will be created, nor how many tokens assigned to the team will be allocated to the advisors.
The project declares that investors will be buying tokens under the SAFT (Simple Agreement for Future Tokens) at the ICO. After the obligatory KYC procedure and emission, investors will receive the tokens owed to them. In our opinion, SAFT is not a guarantee of legal security; there are precedents of SEC claims for projects that conducted token sales in this format.
In general, the essence of the token is understandable, and its existence is justified. However, its economy is described insufficiently transparently. It is unclear whether the pricing of rewards for crowdworkers changes when the token price changes, or how the balance of available orders for labeling and the number of those willing to receive a reward for labeling will influence the pricing. Many parameters for interaction are discussed with the community via Telegram, which we very much appreciate. However, when studying the project’s documentation, many parameters remain unclear. In contrast to the high level of product development, and of the description of the product itself, there is a lack of description of the token economy, which creates additional risks for early investors.
Analysis of Factors Affecting Future Value of the Token
In the process of analyzing the factors affecting the token’s value it is worth noting that the project implements fairly complex functionalities both from a technical and economic point of view.
There are a large number of factors affecting the future price. We consider the most important to be these:
The project’s success and adherence to the roadmap. This could be both positive and negative. Important milestones include the launch and development of the mobile application, the launch of a public beta version of the neural network training, as well as a full-fledged blockchain platform with API support in Q3 2018. Openness to the community and investors in matters of product progress is also important at this stage, and will affect the token price.
The need to purchase tokens for using the platform will lead to demand in the market, which will lead to increase in its rate in leaps and bounds if businesses are to be the main token buyers.
Growth of the platform’s audience. From the point of view of long-term development, we see great potential for using the token on the platform as more and more participants join it. The more users joining the ecosystem, the less any negative impacts on the rate will be, since liquidity will be saturated with a larger number of participants, which will lead to greater turnover of the token and, as a consequence, a decrease in "jumps" in purchases and sales.
With an increase in the project’s audience, more and more unsold tokens will “dwell” with users, since when remuneration tokens are received, not all of them will be immediately sold on the market. This will create cyclical and noncyclic deficits for the token, which will positively influence its future value.
Sales of tokens by ecosystem participants. Crowdworkers who are service providers will want to sell tokens on exchanges after they receive them, without storing them in their wallets. This will be a source of earnings for them, which they will want to spend on other things, becoming participants in an economic system outside the sphere of AI and Dbrain in particular. This will create strong supply in the market and will put pressure on the token price.
Sales of tokens by the team and advisors. The team and advisors will have 30% of the total emission, which is a seriously large and potentially highly influential amount affecting the token rate. The lock-up period is a big plus, but due to this large allocation of tokens to the team and advisors, the risk of a decline in token sales is extremely high in the medium-term.
The popularity and trendiness of the AI market could likely create hype around the service and increase demand for the tokens, which will positively affect the DBR rate.
The factor of platform sales. The platform accumulates a certain number of tokens via the commission charges, which could cause strong depreciation when they are sold on the market to support current activities.
Most factors affecting the price are traditional for utility tokens. Users buy tokens to enter the platform, and sell them for profit and expenditure outside the ecosystem. Ideally, especially with a large turnover, it is enough to create additional motivation for storing tokens in wallets to obtain a 1-5% overhang. This really will lead to effects of "withdrawal from turnover", which will lead to a deficit and to growth with all other parameters being equal.
We highlight the positive effect of the lock-up periods. We do not see acute risks of short-term depreciation; everything will depend on the success of the product being launched, the openness and competence of the team and whether the roadmap is adhered to. Risks increase in the medium-term after a year with the end of the lock-up periods, due to the allocation of 30% of all tokens to the team and advisors.
We see the great potential of the platform in the long term, the stability of the token price in the future with fulfilment of the project’s potential and the absence of serious risks as the project progresses.
Investment Risk Analysis
1) The risks of token price reduction as a result of sales by the team, advisors, and other participants. Since a large number of tokens are concentrated in the hands of potential sellers, we estimate as high the probability of token price reduction in the medium term. It should be noted that the token economy is not well thought out, nor is it fully disclosed; this makes the risks of price reduction substantial.
In addition to this, constant competition between platforms, reductions of the rates of major cryptocurrencies, large-scale sales by token holders are possible factors for DBR price decreases. If the declining market phase takes a long time, the token may no longer be of interest to platform participants as it will not prove to be a reliable settlement or payment means. In this case, potential customers may turn toward competitive offers requiring fiat currency.
2) Risks of a highly competitive market and product development. The competition should not be underestimated. The belief that the Dbrain product has a strong competitive advantage should be shared by other industry participants. Products from competitors are in some ways inferior to Dbrain, but their solutions have been operating in the market for a long time and it will be difficult to catch them up.
3) Liquidity risks. If there is no market-making mechanism or there are no players on either side, this could lead to it not being possible to buy tokens for using the service, or selling tokens for conversion into fiat and subsequent withdrawal. It only takes a few similar situations to cause reputational risk and discredit the token. Perhaps future mechanisms and token reserves will help to manage the liquidity, but with the introduction of a unit of account, there is always the risk of a "dead market" and the inability of one party or another to buy or sell the token at a "fair" price at particular points in time.
4) Risks of delay in product realization. We do not doubt the team’s competence in many areas, including development and management. However, blockchain decisions, especially in the field of artificial intelligence, could be more difficult than it appears at the stage of product realization / development. The current roadmap looks excessively optimistic; we consider the risks for delay in implementing roadmap milestones as high.
5) Legal risks. The project lacks transparency in terms of a registered legal entity, and we note that the use of the Simple Agreement for Future Tokens does not guarantee the absence of legal consequences. In some cases, the SEC may pay even closer attention to this format.