What is an Emailbot?

An e-mailbot is a tool for automatically handling large amounts of incoming e-mails. The bot analyses and processes the content of each e-mail and performs actions based on this information. It can be adapted to suit your needs and industry, and trained via our platform. It is continuously trained and improved via the tagging and language engine (NLU).

What are the benefits of an e-mailbot?

  • Easy to integrate

    Our solution is designed to connect and co-ordinate alongside other software and systems.

  • Highly user-friendly

    We are devoted to ensuring that the interface is as user-friendly as possible. No coding or technical background is required to have an e-mailbot.

  • Unlimited capacity

    The e-mail bot can handle large volumes of incoming e-mails at once, around the clock, without breaks.

  • Increased productivity

    Get things done by allowing repetitive tasks
    to take care of themselves, so you can focus on other areas.

  • Get an overview

    Gain a good and well-organised overview –
    all e-mails are classified completely automatically.

  • We analyse and adapt

    We map your needs for you.
    Please contact us to book a meeting to find out how
    your business can benefit from an e-mailbot.

What can an e-mailbot do?

An e-mailbot processes content and performs actions. You decide what the e-mail bot will analyse and what it should do when. Below is an overview of what it can do.

Forward
The bot forwards e-mails to people or systems

Respond
It can respond to any inquiry

Extract data
Extracts keywords
and information from the content

Connect to API
Can be connected to systems
in order to perform actions

Classify
Categorizes incoming
e-mails by type

Interpret and understand
Analyses your content and learns what is what.

Identify
Identifies with the help of keywords

Filter and organize
Organizes your e-mails and filters out junk e-mail

What happens when the bot receives an e-mail?

To be able to handle the e-mail properly, the e-mail bot analyses the content via multiple layers of understanding:

1. What type of e-mail is it?

When the e-mail bot receives an e-mail, it immediately reads what type of e-mail it is.
Example: Spam, invoice or customer enquiry.

2. To which category does it belong?

The bot identifies the essence of the e-mail, e.g., the subject of the complaint.
It retrieves information via keywords and machine learning,
and understands why the e-mail has been sent.
Examples: Claim, travel expenses, ad sales.

3. What intents does it contain?

We use the word intent to explain an intention or purpose.
What is the intention of the sender? What do the sentences tell us?
Through training, the bot can learn to read the meaning of sentences and determine whether the e-mail has numerous intents.
Example: “I would like to subscribe to your service” “Can I receive an offer?”

What entities are included?

The bot retrieves important information from what we call ‘entities’.
The term ‘entity’ is Latin for ‘that which is’, and is a collective term for ‘unit’.
Examples: City, date, name, numbers, addresses and time.

e-mailbot analysis

How do you get started?

Upload existing e-mails as training data. The bot can then trains itself and learn from existing e-mails in your inbox. It is important to map out the needs – what the bot should be able to understand and what actions you want it to perform.

Step 1

Training via tagging

   To train the language engine, we tag the existing e-mails.

   We tell the bot the type and category of the e-mail, what intents it has and what kinds of entities are specified.

   Once it has been trained with a certain amount of data, the e-mail bot will be able to analyse, interpret and understand the content itself.

emailbot tagging

Step 2

You determine the rules and the action flow

  Choose what the bot will do with the information.

  Set rules and conditions for which actions are to be performed and when.

emailbot action

Integrations

Our solution focuses on the automation of tasks and enquiries, and has been developed specifically for integrations with other systems.

tick  For systems both with and without API.

tick  A comprehensive solution which includes full support and assistance.

For systems in which integrations are not desirable, we will be able to utilise RPA technology Robotic Process Automation to give the email bot access and the ability to perform tasks.

Examples: CRM and case management systems

Language engine

The e-mailbot analyses and makes decisions based on a combination of rules and language engine, also known as Natural Language Understanding.

tick  The bot understands the content and improves as it learns.

Some of our e-mailbots

Sødde for Claims Link

Sødde for the claims settlement company Claims Link

Sødde can determine whether the received e-mail is a notification or a claims case. It recognises the type of case and also considers whether enough information is available to create a case for further processing.

It operates in a common e-mail mailbox along with the other colleagues on the line, and picks, reads, and categorises e-mails. It creates cases, archives files, and also prepares the case folders for further processing.

OSM Martitime

6 digital employees for OSM, a global maritime service provider

We have automated salary payments and provided an e-mail bot, document bot and multiple RPA (Robotic Process Automation) bots that take handle salary payments and customer billing. The digital employee is composed of a group of bots that have been linked together, and functions like a human accountant.

The captain sends sailors’ deductions, overtime hours, advance requests, and the vessel’s expense documents to an e-mail address.

Hire a digital employee

Combine an e-mail bot with a chatbot and a document bot

Our modules can be assembled. When they are combined in this manner, we call them a digital employee.
A digital employee can perform several types of job functions and have different roles.

A digital employee is given access to relevant computer systems and understanding about what it must do. It is then ‘taught’ using relevant data and information, readying it to take its place among its colleagues.

This post is also available in: Norwegian Bokmål

X
X