How our neural network communicated with the FSB

In early August, our OneBox Next neural network entered into correspondence with the FSB and tried to sell itself until the victory, as our sellers behave.

Bypassing all the filters, she managed to register an official appeal and reach the head of the department.



We only found out about this a week ago, when CRM set a task for one of our employees: “Lead #xxxxx sent an answer, the network went into a dead end, you need to continue communication”.
Indeed, on a PDF with a scan, she did not know how to react ...

We ourselves did not expect this!

Two weeks ago, we announced our new development of OneBox Next - CRM with a neural network . The neural network itself is still in the process of development, testing and training, but we already use it in our work:

• it helps our support to make the most suitable answer for the client;
• to our sales, she shows the most successful sales chains that led to transactions;
• and CRM itself collects customer contacts in the field we need and enters into correspondence with them on algorithms that are automatically rebuilt based on previous experience.

PS: It seems that in communications with government bodies only artificial intelligence is able to go through this whole bureaucracy.