ICO date: 16.04.2018 - 22.04.2018

Decentralized machine learning, or DML for short, is a platform that provides opportunities for machine learning and monetization of idle computing power and data.

How it works:

DML leverages idle processing power of billions of devices to achieve new heights in machine learning. This cuts down cost and increases availability to the masses. Machine learning is happening on device with no data extraction, so the privacy of all participants is protected.

Using blockchain technology makes the whole process truly decentralized and transfers the power of Big Data from governments and corporations to the individuals.

DML also acts as a marketplace for machine learning algorithms. Developers can monetize their skills by offering machine learning applications on the marketplace and get rewarded by DML tokens. Same goes for data owners, who can make use of the idle processing power they possess, using software that runs on the background, or raw data sets. The marketplace also has space for machine learning specialists who can offer to fine-tune the algorithm to perform the tasks needed by a business client.

All deals are secure, as they are held using smart contracts. Smart contracts also make it possible for users to distribute rewards among all participants of the deal, the data owners, the nodes in the protocol, the developers and the machine learning specialists.

Customers of the platform can reduce the bar for applying machine learning in their business drastically and get better prediction results. They can develop their own solutions or purchase already available ones, listed on the DML marketplace. Users can choose from a wide array of types of ML algorithms sorted according to a business they can be best used in, and fine-tune them even further to serve their exact purpose.

Prediction results get even more accurate because DML uses a federated node to aggregate prediction results sent by other nodes. DML implements crowd fine-tuning of the algorithms to constantly improve the quality of services it provides.


DML brings machine learning to the mass market. More so, it significantly improves on machine learning protocols and makes it truly decentralized. The fact that they emphasize the constant necessity for improvement of their services and involve a developers community in the crowd creation of the algorithms means that there is much innovation ahead.

Innovation score: High

Token value:

DML token is the platform’s utility token. It is used to make payments to the participants of machine learning activity via a smart contract. With the platform’s rise in popularity, the amount of transactions will increase, driving the demand for the token. ICO participants can cash in on their investment when the price of DML goes up.

Token value score: Medium

White paper:

DML’s white paper is rich in information and contains flow charts that explain the difficult pieces. It explains the market potential the project has and talks about key features of the marketplace and the platform’s functioning in well structure brief and understandable way. The white paper also describes benefits that different participants of the marketplace can get and outlines future possibilities for the protocol evolvement.


The only drawback that it does not contain a business model description and a marketing plan, but the DML team gladly provided us with that information.

White paper score: High


The company has already released a prototype of DML Algo Marketplace in February 2018, they plan on launching beta in April. The marketplace will be fully launched in May. The alpha version of the next protocol, the one for on-device decentralized machine learning will be launched in July 2018 and will be online in September. In December of the same year the team will be working on increasing DML scalability.

The third protocol necessary for the platform’s functioning will be fully released by the second quarter of 2019, the fourth by October 2019. Final stages of work on the project will be finished in the third quarter of 2020.

Roadmap score: Medium

Business model:

The research DML provides in their white paper indicates that the revenue from Big Data analytics will grow up to $210 bln by 2021, the market worth being $2 trillion. So the company taps into a very profitable industry.

The team does not intent to have a listing fee in first three years of operation to attract more users. Transaction fees will be determined by the market itself and will go to the maintenance of the platform, along with a part of ICO funds reserved for that.

Early users will get an ecosystem bonus, a payment in DML tokens to encourage new contributors to join in. The bonus will decrease over time, so it will be really profitable to be among the first to download the software and rent out your idle processing power. The company also partnered up with a platform that can be the source of early demand for machine learning services.

Though DML is targeting all over the world, they are starting from San Francisco, Moscow, Hong Kong, Singapore and Taiwan as they have connections with machine learning community there. The company will launch hackathons to source machine learning developers.

Business model score: High


Project lead director is a founder at two tech award winning companies, a growth lead in early stage startups and a veteran digital marketer. One of their software and blockchain developers is a former software developer at Uber and at a company later acquired by Goldman Sachs. DML’s security officer has over 20 years of experience in IT.

The advisory board has Big Data, blockchain, PR, marketing and business advisors. Machine learning advisors are plenty. The business advisor used to work for Disney and EA Sports.

Team score: High


There are already some giants like Google Cloud and Microsoft Azure working in the industry, but they are hardly any competition to DML because they are centralized and very expensive.

Indeed it is hard to find a platform that would secure your data by blockchain and encourage crowd fine-tuning of the algorithms, making the whole thing very accessible, even though there are a lot of players on this young market already. DML has what it takes to take on the competition.

Competitors score: High

Overall thoughts:

DML enters a market that is predicted to grow significantly. It has a fresh idea that will attract a lot of users and a dedicated team with connections within the machine learning community. Investing in DML tokens is highly likely to bring a lot of revenue.

Whitepaper: Whitepaper

Overall Score


4.3 / 5

Risk score




ICO start date: 16.04.2018
ICO end date: 22.04.2018
Hard cap: 28,000 ETH


Ticker: DML
Token price in ETH: 0.00026


ICO Platform: No
Registration Country: Singapore