How autonomous agents use machine learning to transact on the network

Dec 16, 2019

  • agents interact to share knowledge, negotiate and trade autonomously
  • Users of the network benefit as they are provided with the information to optimize their time and resources more effectively

One of the aspects that makes unique is the capabilities of our Autonomous Economic Agents (AEAs), or agents for short. These are not just bits of software that represent data — they work autonomously on your behalf and conduct economic transactions to make your life better. In this article, we’re going to outline exactly how they do this.

On the network, individuals or companies with data, represented by their agents, connect with agents of individuals or companies that seek it. Agents operate on the Open Economic Framework (OEF). This acts as a search and discovery mechanism where agents that represent data sources can advertise the data they have access to. Likewise, individuals or companies looking for data can use the OEF to search for agents that have access to the relevant data.

Imagine a company is seeking data to train a prediction model. When the company’s agent has connected to the agent representing the data source, it will ask it for information regarding trade terms. The agent working on behalf of the data provider would then provide the terms under which it is willing to sell the data. The agent selling access to the data is likely to seek the highest price possible, while the agent buying access to the data would like to pay the lowest possible price. However, the agent selling the data knows that if it charges too high a price, it will miss out on the transaction. This is because the agent seeking the data will not accept the terms and will instead attempt to buy the data from another source on the network. If the purchasing agent does find the terms acceptable, then it will pay the agreed price to the selling agent via a transaction on the ledger. After receiving the payment, the agent selling the data would send the encrypted data across the network.

Apart from the initial set up, the whole process is fully automated and performed by agents. This means company employees are able to work without disruption, while the prediction model accumulates relevant, anonymised data. By obtaining the data, the company buying the information is able to more effectively train its model, which it can then use to make more accurate predictions. Such predictions could be used across any industry. For example, if a company has collected data regarding the maintenance of lorry components, it could predict when it is most likely that parts will be in need of repair. This type of information would be useful to couriers and businesses working in logistics. It would enable companies to bring the relevant vehicles in to be checked by mechanics before they broke down and it would minimize the risk of the supply chain being disrupted by the vehicle breaking down while out on the road.

Get Involved