Category

  • Building bots for chat, SMS, email, voice and more char

    Connecting Emails, SMS, and Voice to Workflows

    Email Bot

    As a first step, customers should determine if their email system supports API for Vizru workflows to access email content.  If not, they can simply write a forward rule to a parsing service, such as Mailparser or Mailchimp.  Once an email content is forwarded, Vizru workflows can parse and categorize its recipients, sender, subject, email body, and attachment.  Once parsed, the necessary data is sequenced and sent to the NLP engine to identify the email intent.  Now, Vizru Email Bot can take a series of actions based on the email content and its intent such as auto-reply to emails, OCR and file documents, forward, approve, create and escalate tickets, and much more.

     

    SMS Bot

    Vizru works with top SMS gateway services such as Twilio, Nexmo, and Retarus to enable two-way SMS requests and SMS-based transactions.  Similar to the case of chatbot, customers can repurpose the chat journey for SMS as well as chat.  The workflow can be designed to strip out the rich UI of chat and just send users content in plain text.  Vizru session management handles SMS flow a little differently from chat as follows:

    • Step 1:  Vizru Web application -> Session Created -> Workflow invokes SMS Gateway -> Workflow sends update -> User receives update SMS on phone
    • User sends response -> SMS Gateway receives -> Post message to workflow -> Session Updated -> Workflow parses data -> Workflow sends to NLP -> Workflow saves data to Mongo -> Workflow prepares the next step -> Repeat Step 1

    Use Case:  Payment acceptance, multi-factor authentication, user verification, appointment confirmation, and more.

     

    Voice Bot 

    Vizru Gateway can interface with voice assistants like Alexa or Google Assistant, and also to VOIP services such as Vonage through their API.  Vizru workflows then translate voice to text and follow steps similar to that of a chatbot.  Once tasks are completed, the workflow notifies the Gateway which in turn once again taps into voice assistant services to convert text to voice or optionally leave voice messages automatically. 

    Note:  The accuracy of this service depends on the clarity of the sound, recording, and accent of the speaker.