title | description | coverImage | layout |
---|---|---|---|
Hornbill AI Configuration |
This guide covers the configuration and implementation of HAi (Hornbill Ai) features within your environment. The document outlines step-by-step instructions on enabling HAi capabilities for your user account, specifying the necessary roles, and accessing the necessary settings in Hornbill. |
/_books/servicemanager-config/administration/images/hai_cover.jpg |
article-toc |
::: important HAi is currently in a closed beta, speak to customer success should you want to take part. :::
All HAi features are disabled by default and must be individually enabled before they can be used by end users. To enable any of the HAi capabilities, your User Account must be associated with one of the following roles:
Role | Description |
---|---|
Service Desk Admin | This role is for a Service Desk Administrator. This includes, an elevated visibility to Requests and associated actions, the ability to configure Service Manager features via the Administration Tool and the option to restart a failed BPM within a Request. |
Additionally, any user with the application right rightA.administerServiceDesk can enable HAi features.
- Log into Hornbill with the relevant access role.
- Click on the settings gear icon in the lower left of the screen.
- Change the dropdown from '''My Personal Settings''' to '''Service Manager'''.
- Click on HAi under Administration in the left-hand menu.
- Enable the required feature(s).
To access the HAi capability of Hornbill Service Manager, your User Account must have one of the following roles associated.
Role | Description |
---|---|
HAi User | This role allows users to access HAi functionality inside of Service Manager |
By enabling any of these optional HAi features, you are agreeing to allow a third party to process your data (OpenAI), Data from prompts and underlying request data is passed to OpenAI under their business terms of use and their enterprise privacy policy. This prevents usage data from being used for training new models, the input and output of prompts is currently stored by OpenAI as outlined in their Documentation, all data is sent using encrypted connection.
The following areas of functionality send request data during each invocation and are document as follows:
When summarizing a request the following data from the request you are in is processed:
h_itsm_requests.h_description
h_itsm_requests.h_summary
h_itsm_requests.h_fk_user_name
The timeline of the request is filtered down with the following filters '["Update","Email",'Customer']'
Each post in the timelines sends the follownig
actorInfo.name
content
Each comment associated to any filtered posts in the timelines sends the follownig
actorInfo.name
comment
Custom Questions (first 100) are passed, excluding the type file-upload
and label
and are paired with the question text as follows:
h_question
h_answer_value
When suggesting a resolution against a request the following data from the request you are in is processed:
h_itsm_requests.h_description
h_itsm_requests.h_summary
h_itsm_requests.h_fk_user_name
The timeline of the request is filtered down with the following filters '["Update","Email",'Customer']'
Each post in the timelines sends the following
actorInfo.name
content
Each comment associated to any filtered posts in the timelines sends the following
actorInfo.name
comment
When generating a knowedge draft from a resolved or closed request the following data is sent:
h_itsm_requests.h_description
h_itsm_requests.h_summary
h_itsm_requests.h_resolution
Text assist combined with snippets or used withing a workflow and passing in variables can pass any data a user has selected to pass to OpenAI, nothing is sent automatically. In the case of snippets, when the snippet is selected the analayst will see the data before the text is passed as part of the prompt when using Text Assist after a snippet.
Usage of the OpenAI API's goes through an API Key provided by hornbill. Currently there are no usage limits outside the limits provided by OpenAI. It is possible to override the API to use an API Key of your organization's choice by overriding the system setting integration.openai.apiKey.
Further details on this here: https://docs.hornbill.com/esp-fundamentals/productivity/ai-assist