Product Release Notes Q2 2022

Models can be trained in-app by users without the help of the YM ML team
Ability to re-activate prior versions of a path, or to convert them to a draft version
A generic retriever that can accept any data format and initiate a trip
Hubspot retriever utilizing the generic retriever
Added a new endpoint for retrieving usage statistics related to billing and model training
Added additional information about “first message” to the Front retriever
Switched to human friendly names in the data output pills shown in the Builder
Added “Reset Password” link to login page
Greatly improved copy/paste functionality for expressions
Allow for data pills in expression to be easily replaced
Added requirement for users to change password on initial login
AI Model Security Control Improvements
Allow data to be uploaded directly into the app for AI model training
Added “files” as a source in the Labeling tool to support labeling of uploaded data.
Expanded trip metadata for several retrievers to provide additional system-specific data for use in trips
In app testing of simple PyMeadows expressions and the ability to edit same directly from a step input
Exposed step descriptions in the Journey Builder
Added experimental application capability within the application. First experimental application is for clustering data to expose hidden or unlabeled intents
Added tooltips to Data Pills to allow identification of the parent step

A tool for building single, simple API calls with minimal coding was created
Notes step for the Hiver app
Slack integrations
- Bot to enable notifications from Slack Forms
- Make a comment in a Slack channel
- Reply to a comment in a Slack channel
Redirect processing to new path without changing intent
Lookup package tracking information for various carriers
Front - Get messages in conversation
Gmail - Threaded Draft Reply
Hubspot - Create Ticket
Hubspot - Find Customer in List
Hubspot - Get Associated Deals
Hubspot - Get Company for Contact
Hubspot - Get Ticket Information
Hubspot - Link Deal to Ticket
Hubspot - Lookup Company
Hubspot - Lookup Deal ID with List of Addresses
Jira - Move tickets to another project
Jira Add Comment to Jira Issue
Jira Close Issue in Jira
Jira Create Issue in Jira
Jira Link Issue in Jira
Jira SM - Edit Ticket Fields
Okta - Get user information
Okta - Get user information
Okta - initiate resend activation email to user
Okta - initiate reset of user password via sending link
Okta - Return list of assigned applications
Okta - Search for User by email
Salesforce - add Contact to a Case
Salesforce - link Account to a Case
Tools - Check Lock File
Tools - Convert File to PDF
Tools - Create a List
Tools - Create Lock File
Tools - Output Lines from A File
Tools - Plain Text Search/Replace in text
Tools - Unlock File
USPS - Get shipping status using tracking #
Zendesk Edit Ticket Fields include Custom fields

Maximum retries for each step was added as a configuration for Zeebe BPMN
Validation for docket image name was added to avoid problems with typos in factory version configurations
Trips occurring during updates to nodes no longer fail
Improvements to environment auto-scaling
Fixed an issue where a high volume of trips would delay the availability of data in the Map
Improved performance of the data redaction routines
Resolved a bug where output from previous steps was sometimes sent to the current step when it was not requested.
PyMeadows expressions were enhanced to allow for multi-line expressions to be used
Resolved Out Of Memory errors in the backend
Improved the security of backups
Made is possible for a path to be published without interrupting the processing of an in-flight trip on that same path
Updated codebase so all error types have a unique "errorCode"
Added feature flagging to enable management of new features or rollout to individual namespaces
Fixed bug that would cause errors when SSL certificates were pasted into connections

Last updated 1 year ago