DRYiCE Lucy v7.0

The release notes capture the details of features and functionalities introduced in DRYiCE Lucy as part of its major release version 7.0. This release follows Assisted upgrade, details on which can be found in the Upgrading to new version section. For any queries/clarifications, please reach out to Support.

What’s new?

Following are the newly launched features and functionalities as part of DRYiCE Lucy v7.0

• Lucy on GCP Cloud: DRYiCE Lucy is now available on Google Cloud Platform (GCP), while leveraging the capabilities of Google Dialogflow instead of IBM Watson. The following consoles are available in Cognitive Console for the current release:

• Enterprise Connectors: New out-of-the-box connectors are available for enterprise applications:

• Natural Language Understanding (NLU): For enhancing Natural Language capabilities, added feature for understanding and extracting Natural Language feature sets that are associated with Emotions, Sentiments, Semantic roles, Critical Keywords, Categories, Linguistic Syntax, and Concepts. Each feature set is configurable and is also supported in multi-lingual.
• Sticky Sessions (Cross-channel Continuum): This feature allows users to maintain sessions across all the channels (for example, Web Channel, MS Teams, Slack, SMS, Mobile Apps, etc.) that are enabled on the bot. Lucy will now be able to handle discontinued conversations in a channel from another channel, so that users don’t need to restart conversations from the beginning in case communication channel is switched or chat is abandoned.
• Configurable Feature Sets: This feature allows users to configure NLP feature sets for tenant (depending on the requirements of clients). As a result, these features are available in the UCD console while configuring use cases. It allows users to control relevant metadata extraction from utterances, such as concepts, entities, keywords, categories, sentiment, emotion, relations, etc
• Entity Annotation: This feature allows finding contextual entities automatically with multi-word selection and annotating them in the UI. Adding annotation helps NLP to understand even untrained information on picking objects and references.
• Navigation Assistant (Tour Guide): This feature allows new users to familiarize Chat UI on the Web. It guides all functionalities that can be used by the user while accessing Lucy on the web. All messages in Navigation Assistant are configurable as per the customer’s requirement. Additionally, this feature can be enabled or disabled by the admin from the Master configuration of the tenant.
• Configurable Chat History: This feature allows tenant admin (SMEs) to customize the number of days for which chat history can be displayed on the Lucy’s web interface.
• Did you Mean Helper for Microsoft Queries: This functionality is used for handling fallbacks to search through Microsoft official documentation. Users can find relevant answers when the bot is unable to provide the correct response. This functionality is supported across all channels.
• Right to Delete Console: This feature is added in compliance with GDPR policies, where users can request data erasure. This feature redacts user-related data from the system which includes the removal of the messages that they have exchanged.
• SOP Translator: A new console is added to Cognitive Console for the Translation of SOPs.This console allows users to add translations to the contents of a newly created or already existing SOP (Standard Operating Procedure).
• ITSM Catalog Trainer: This feature is used to synchronize the service catalog for ITSM applications across multiple channels (e.g.: ServiceNow, Cherwell etc.). It allows quick import and use case creation / training of a service catalog from any ITSM application (ServiceNow, Cherwell, etc.). This can be done by simply importing an Excel (XLS) file into Lucy instance.
• Task Modules: This functionality allows custom-built webpages or media components like videos to be displayed in MS Teams channel.

What’s enhanced?

Following are some of the existing features and functionalities that have been improved upon as part of DRYiCE v7.0:

• Enhanced Intent Recognition: This functionality allows swiftly switching between a newer model which provides better Intent classification, with minimal number of variations added to the intents.
• Multi-lingual support for DRYiCE Optibot Livestation: For enhancing out of the box Optibot Livestation capabilities, a multi-lingual feature is added in the current release. It allows user to connect with Agent and communicate through their selected language.
• Skill Deletion: Provided an option for skill deletion to Admin (SMEs) of tenant from UCD Console. Now, Admin (SMEs) can delete any unused or stale skill directly from UCD Console without raising any support ticket for it.
• Smart Email Channel Signature Removal: This functionality effectively removes email signatures when used as email bot. This can take any signature format which includes the address or phone numbers. Currently, this functionality best works for the English language.
• Email Channel – Support for buttons: Simple interaction buttons so that users can click to respond as an email to the bot back instead of typing.
• Adaptive card Carousal:Support for Carousal-based adaptive cards which would greatly increase the UI capability to provide a better experience with more interactive content management for MS Teams and MS-partnered chat windows.
• OTB Rule for Knowledge Searches: One-click OTB Rule to configure knowledge searches to any channel including Teams and other non-web channels.

Following are the bugs fixed as part of the DRYiCE Lucy v7.0 release:

• Fixed: Use Case Design Console UI
• Fixed: Removed Keyboard pop-up whenever Lucy responds while accessing through the mobile browser or mobile app
• Fixed: UI Caching issues on page load
• Fixed: Inconsistency in chat attributes storage
• Fixed: UI fix on zooming of images
• Fixed: Adaptive card JSON padding issue
• Fixed: Document fields on Knowledge Management Search (KMS)
• Fixed: Lucy User Management System (LUMS) login error prompts
• Fixed: Context poisoning on agent transfer
• Fixed: SOP backup-restore

The following changes were done which in turn improved the performance and security of DRYiCE Lucy v7.0 in comparison to the older versions:

• Implemented Brotli compressions on the resources loaded on Global Admin and Tenant Admin consoles
• Faster load times and no server strain on loading assets
• Adjusted function performance to allow more load
• Made Changes in Route management
• Node packages upgraded for security and internally used libraries to obtain better performance
• Optimized memory utilization by reducing memory leaks including front end
• Optimized socket events for consumption of lesser memory and improved speed
• Faster Service worker deployments
• Zero Server load for page load times on admin consoles
• HTML and UI element restructuring for adhering to the latest browser standards
• PWA specific changes on UI of Admin dashboards
• Under-the-hood changes for removing cyclic dependencies to improve reliability
• Pushed to the Latest Active LTS version of NodeJS (v16)
• Channel registration optimized to pick for specific versions

Depreciated Feature means that the feature is at the end of support and its alternative must be used. Although the feature will continue to work, it is not recommended to use the below feature:

• None

Following are the limitations/restrictions with respect to the features and functionalities released as part of DRYiCE Lucy v7.0:

• Google Cloud Platform (GCP) Variant is equivalent to Lucy’s v6.1 release
• Tenant onboarding is not supported for GCP-specific tenants on Global Admin
• Scheduler Module and UCD consoles are still unsupported on GCP
• Multi-channel not supported on GCP
• Other Limitations not listed here from the previous versions persist

DRYiCE Lucy is hosted on IBM Cloud & Google Cloud Platform and utilizes PaaS services. System Requirements are Not Applicable (NA).

DRYiCE Lucy v7.0 release is a major version release that includes Assisted Upgrade. DRYiCE Lucy follows a zero-downtime deployment strategy with automatic rollback in case of failure of deployment for any feature/ functionality during the upgrade.

To upgrade to DRYiCE Lucy v7.0, the Product Professional Services Team follows the below steps:

Note: The steps mentioned below are undertaken separately for each tenant within a customer environment. This means that if there are multiple environments like Production, Development, and Staging/User Acceptance Testing (UAT), each of these environments follows the below instructions:

1. The PS Team verifies the current version of DRYiCE Lucy implemented in the customer environment (DRYiCE Lucy v7.0 has a “One-click” automated upgrade functionality to upgrade a selected tenant. For DRYiCE Lucy versions older than v7.0, the upgradation process is manual. For this, a mutually agreed time slot is confirmed between the customer and the PS Team).
2. The Providers will access and review the existing functionalities/ feature within the customer environment and share the impact with respect to upgrading to v7.0.
3. Non-Production environments should be a logical replica of Production before upgradation.
4. This upgrade is through a Provider-owned model, where providers find the upgrade card for the tenants who are eligible or not through which the tenant can be upgraded. This can be found on the Configure page of the tenant by clicking the gear icon next to the tenant on the tenant listing page.
5. For Provider-owned Model, the Provider Admin or the Provider SPOC can follow the steps mentioned on the Upgrade page for upgrading the tenant automatically.
6. In case of any issue, please raise a support ticket on the Support link of DRYiCE Lucy. Lucy Professional Services team also needs to create a task on the JIRA portal.

DRYiCE Lucy is a conversation AI platform. It provides extensive capabilities to create conversational agents which mimic human interaction, learn, and adapt to user needs through smart conversations by leveraging enterprise-grade Natural Language Processing (NLP) and Machine Learning (ML). Users can leverage Lucy to find the relevant information available across different enterprise systems.

Lucy helps organizations move up the maturity curve and enhance the user experience for employees and customers. By leveraging advanced natural language processing Lucy reduces human errors and increases productivity. It comes with out-of-the-box use cases for various scenarios and can be easily extended to cover the “Cognitive chatbot” needs of modern enterprises.

For any queries/clarifications, please reach out to us on Support.