How does AI-powered email management work?

Extracting meaning from emails

  • Natural-language processing to extract the essence of the message
  • Identify customer’s issues when using the product / service
  • Identify actions / tasks to be taken

    • Text requests understanding, multi-intents detection and NER (Named Entity Recognition), to be used by the customer services and support teams.


Gathering additional information

Eliminating manual check for internal and external additional info:

  • Existing customer in the database? Order status?
  • Use contextual data to automate final decision


Generating answers and tasks

  • Natural-language processing to extract the essence of the message
  • Identify customer’s issues when using the product / service
  • Identify actions / tasks to be taken
EmailTree understands the content and automatically generates reports to help customer service and support team.

Learn and take actions

Learn from every action taken by human operator. Some actions are already taken by EmailTree and emails are sent out. Other actions need review by human operator before sending out the email. Next time, in this scenario, EmailTree will automatically send out the proper email.


Completely task automation

EmailTree learns from the previous decisions and automatically:

  • Generates actions
  • Prepares customised offers
  • Proposes solutions for technical issues (Support Level 1)
  • Escalates to Support Level 2 or 3
  • Generates a confidence level, based on the previous steps
  • Composes the email content and send emails
    Human operator may interfere or not to change elements in the email to be sent.

Learning score improvement

Each decision contribute to the learning process. Day by day, the confidence level improves and EmailTree performs better.