Thank you for installing Koji, the сhatbot for Jira Service Management! It automates your service desk and helps users with everyday tasks.

This page describes the first steps of getting familiar with Koji and its functionalities.

Backoffice

All right, Koji is ready to go. Let’s go to to the backoffice to see what you can do now.

The backoffice is the place for managing Koji. You can fine-tune the behavior of your chatbot, study analytics, orchestrate users and their roles, and even train Koji.

To open the backoffice:

1. Switch to the corresponding project in your Jira environment (click Projects at the top of the window).

2. Click Koji backoffice in the navigation menu on the left side of the window.

Open

Setting the configuration

First of all you might want to check the settings that define your Koji’s behavior. To do so, click Koji settings in the backoffice to see and modify settings.

Open

Publishing the chat

By default the chat is not displayed in your portal. Once you have checked the settings and made sure all systems go, you can publish the chat, so people can start communicating with Koji.

Getting to know users

Probably your Jira project is not empty, and you have a number of users there. Koji does not know these people yet. It get familiar with those users, who open the portal with enabled chat. They are included into Koji’s list of users.

Helping Koji become more efficient

Once users begin talking with Koji, it collects the conversation details. These are crucial for understanding users and making Koji more efficient.

Koji is powered by artificial intelligence, which means it can be trained. It can become smarter day by day, but it will need your help.

• FAQ allows you to create articles and build a knowledge base as a result.

• Conversation training allows you to train existing conversations to make them more efficient.

• Knowledge training allows you to enrich the list of users' questions associated with a given response.

Creating FAQ

Create an internal knowledge base that provides users with short answers — FAQ.

Conversation training

Analyze real-life conversations and improve the bot’s ability to provide useful replies.

Knowledge training

Review the articles that users have marked “useful” or “not useful” in the chat, and train Koji. A well-trained assistant offers useful articles to users when appropriate, speeding up the search.

Studying analytics

Koji monitors your environment’s metrics, from conversations and knowledge bases to external interactions and license management.

Click Analytics to view details and better understand your users and analyze how well Koji is performing.