AI is the go-to technology for today’s industry which inevitably benefits everyone and makes processes even more efficient and smarter.
The application of AI can be seen in practical problems in real life where extremely impressive results have been achieved especially in technologies that relate to vision, speech, machine translation, as well as decision making.
Even if you move out of the scope of research and development, AI is starting to gain a foothold in many businesses’ processes, with important processes such as database systems and user queries being powered by AI.
However, there are inherent problems within the industry that needs to be dealt with before it grows even bigger.
The learning process of AI and other machine learning models are hindered by an overload of data that are unlabeled and uncategorized.
These data can be classified but converting random dataset to a labeled dataset (a set of data with classes or tags to inform or educate the machine learning model) is very expensive and time-consuming.
Because of this problem as well as differences in businesses’ resources and processes, the gap between early AI adopters and individuals who adopt it much later will drastically widen, which grants early adopters an unfair advantage over others.
The main focus of the Neuromation platform will be on synthetic datasets that have been proven time and time to yield compelling results for use cases.
The use synthetic datasets in machine learning and other processes will significantly decrease the cost of AI adoption as well as ease the process of mass adoption by the public.
So, what does the platform really do?
The platform will utilize a smart combination of datasets, decentralized computing power, as well as advanced machine learning models to help AI researchers across the globe connect with each other.
In layman’s terms, Neuromation aims to build the world’s most advanced AI data platform where businesses can access the platform and library of datasets to utilize and adopt AI technologies for their own use.
There will be an integrated exchange within the platform where users can either contribute or purchase the components of an AI model with one another; much like buying separate parts of a Lego set to build one complete project.
The Neuromation platform moves one step further ahead efficiency-wise by using distributed computing as well as proof of work tokens to help with the platform’s processes and transactions.
With offices in San Francisco, Tel-Aviv, Tallinn, Kiev and Moscow, Neuromation is led by Maxim Prasolov, a respected serial entrepreneur with tons of excellent projects and businesses under his belt.
Another important individual in the platform is Sergei Nikolenko, a respected CRO and scientist in the industry.
Sergei is a key individual in the machine learning department and contributes greatly to the platform’s technological growth while being the platform’s researcher in the field of machine learning and analysis of algorithms.
Neuromation also has two key advisors to the platform’s goals; they are David Orban, the founder of Network Society Ventures, and Andrew Rabinovich, the director of deep learning at Magic Leap.
Orban is responsible for the platform’s international integration into a deep learning community. Andrew, on the other hand, is the main man behind Neuromation’s object recognition technology, allowing the platform to achieve unprecedented results in algorithms training based on synthetic datasets.
Prasolov will lead a team of experienced developers and talented individuals to help power Neuromation, a platform set to revolutionize the AI industry.
As mentioned earlier, Neuromation aims to combine the components needed to as well as synthetic data to build efficient deep learning solutions—all of this to be done on one platform.
Service providers registered on the platform can provide specific resources for the execution and development of the three key components: synthetic data sets, distributed computing services, as well as Neuromation’s machine learning models (which will be explained further down below).
Here is an excerpt taken from Neuromation’s whitepaper that describes how the platform will work:
“Imagine a place where you can go and easily address all requests to acquire AI capability. A vendor will create the data generator for you, then a group of Neuromation Nodes will use the generator to quickly create a massive virtual data set. You can then select a set of Deep Learning architectures to train on that data. Then another group of Neuromation Nodes will do the training in record time!”
Neuromation will consist of three core modules, each with its own processes and libraries.
Synthetic dataset module
Users have access to the synthetic data set module which can be utilized to create data generators, order data sets from the data generator, as well as requesting data labeling for models.
Libraries available in this dataset can be applied to deep learning models and data generators, as well as a library of datasets which can be acquired from the platform’s integrated marketplace.
Machine learning module
The machine learning module will include deep learning model processes, importing models to the platform, ordering training on selected data sets, as well as acquiring custom models from the marketplace.
Like the synthetic dataset module, this module will include deep learning models which can then be offered on Neuromation’s marketplace.
Users can purchase tokens in the user module as well as having access to the platform’s services and the ability to use or provide processing power to the network.
The user module will contain the user data and user models library.
The back-end of the Neuromation platform will also contain the Market Module which enables the efficient matching of buying and selling orders for datasets, models, and labeling services as well as enabling liquidity in the system.
NeuroTokens (NTK) (which will be sold during the token sale) are the primary exchange mechanism for synthetic data generation, distributed model training, data labeling and other AI services in the Neuromation platform.
Neuromation also applies a token burning policy which is the process of buying back tokens from the open market and removing it from the total supply to increase the inherent value of the tokens.
Here are the details of the upcoming NTK token sale:
Token name: NTK
Token base: Ethereum (ERC-20)
Token supply: 100,000,000
Token pre-sale duration: currently live till 1st of January 2018 (pre-sale duration extended)
Token sale duration: January 7th – 15th of February 2018
Token sale target: 60,000 ETH (hard cap)
Token exchange rate: 0.001 ETH = 1 NTK