Several words about our implementation of the masternodes:

The main purpose of masternodes are:

 * they will be used to perform ML tasks and acts as workers in distributed tensorflow environment
 * these tasks divided in two parts;

    * object detection of varroa destructor using images provided by beekeepers
    * model training to increase quality of detection

 * also masternodes will acts as "ethereum-parity" node to increase network connectivity

The rewards will be distributed according to the amount of performed work, so machines with GPU are very appriciate

We use well known technologies and languages like Erlang/OTP, Python, Docker, ...

Also according to the best user experience, our masternodes:

    * able to perform self configuration
    * able to perform self upgrade, so no need perform upgrade by user