Soul Machines Studio Release Notes
- Brook Warner
- Jon Borromeo (Deactivated)
- Lia Pool
v2.61 | 05 August 2024
UPDATES
We are continuing to improve turn-taking with the Digital Person. In this release, we have improved the Digital Person’s ability to detect that it was interrupted and give the participant extra time to complete their turn. When the Digital Person is speaking and is interrupted by the user, the Digital Person will wait a half-second longer before speaking next time. A half-second is added each time the Digital Person is interrupted up to a maximum of three seconds. Each time the Digital Person is not interrupted while speaking a half second is removed from the wait time. You can enable this feature by selecting “Enable users to verbally interrupt the Digital Person” in your project configuration.
When Interruption and Streamed responses are enabled the Digital Person knows where it was interrupted and can continue from that point if asked.
Streamed responses are now out of beta! New projects will use this by default. You can opt-out by toggling this off in the Generative Conversation Base settings.
BUG FIXES
We fixed an issue where “Boost Expressiveness” was causing Digital Person speech to be cut short in both live deployments and in the video creator.
v2.60 | 22 July 2024
UPDATES
We are continuing to improve turn-taking with the Digital Person. In this release, we have improved the Digital Person’s ability to recover from verbal interruptions by adjusting the conversation history so the LLM understands it was interrupted.
BUG FIXES
We fixed an issue where typing into the chat box wouldn’t interrupt the Digital Person.
We fixed an issue where adding
@hidecards()would create an empty speech bubble in the chat transcript.We fixed an issue where empty speech bubbles were created in the chat transcript when the Digital Person was interrupted.
v2.59 | 8 July 2024
UPDATES
We’re updating how we handle billing and eliminating the need to sign in to the billing portal a second time. Users should now be able to manage their billing via the “Account & Team” page. We are rolling this feature out gradually.
We updated the content card skill to enable you to show the cards as the Digital Person speaks, rather than all at the start.
We added a Spanish Language coach to the AI Assistant Gallery on http://soulmachines.com
We made it more clear when Alex the Language Coach was in ‘tutor’ mode or ‘roleplay’ mode, by having the camera change.
BUG FIXES
We fixed an issue when the Digital Person would speak punctuation marks.
v2.58 | 25 June 2024
UPDATES
We’re updating how we handle billing and eliminating the need to sign in to the billing portal a second time. Users should now be able to manage their billing via the “Account & Team” page. We are rolling this feature out gradually.
We’re continuing to improve the experience related to turn-taking. This release, we’ve improved the naturalness of how the Digital Person’s speech is finished when interrupted by the user.
BUG FIXES
We fixed an issue for generative conversation streaming [beta] which caused the Digital Person to prefer replies of one sentence in longer conversations
Release v2.57 | 11 June 2024
FEATURE
We’ve improved the Digital Person’s turn-taking ability, resulting in fewer instances of the Digital Person speaking over the end-user. You can toggle this feature on and off per project by selecting “allow the user to verbally interrupt the Digital Person”. Please note this feature is in beta. Read more here
UPDATES
We’re updating how we handle billing and eliminating the need to sign in to the billing portal a second time. We are rolling this feature out gradually.
We’ve increased the amount of conversation history that a Digital Person can recall
We enabled a 14 day trial for the Pro plan
We updated the syntax for the Pronunciation skill to use a
|instead of a-. Read more here
BUG FIXES
We fixed an issue where the transcript didn’t reflect the spoken content when using Generative Conversation
We fixed an issue in Dutch language which caused the Digital Person to exit after speaking the word "eerste"
Release v2.56 | 28 May 2024
FEATURE
We now support Deepgram for speech recognition. Deepgram offers excellent speech recognition abilities and boasts low latency.
We released a “Pronunciation” skill, which allows you to teach your Digital Person how to pronounce certain phrases phonetically.
Coming Soon: We’re improving the Digital Person’s turn-taking ability, resulting in fewer instances of the Digital Person speaking over the end-user. Once released, you can toggle this feature on and off per project.
UPDATES
The Voiceflow Skill now allows you to set your “Base URL”.
We’re updating how we handle billing and eliminating the need to sign in to the billing portal a second time. We are rolling this feature out gradually.
We’ve added new Google and Microsoft voices, and removed any that were no longer supported
BUG FIXES
We fixed an incompatibility between the Watson base conversation Skill and the Generative Conversation Fallback Skill.
We fixed a bug where updating your Organization name didn’t reflect in the interface.
Release v2.55 | 29 April 2024
UPDATES
We’ve made several updates to the Generative Conversation Base & Fallback.
Under “Generative Conversation Base” there a new toggle to match the conversation style to the behavior style chosen during set-up, which will be toggled off by default. When turned on, the chosen behavior style tags will be added to the experience.
We’ve renamed the “personality” field in the set-up to “Additional conversation style prompting.” Add more personality queues here to enhance your experience.
BUG FIXES
We’ve fixed a bug so when using the Generative Conversation Base & Fallback the Digital Person doesn’t add any interruptions by the user to the chat transcript. Additionally, all spoken greetings by the Digital Person are now included in the transcript.
Release v2.54 | 15 April 2024
UPDATES
The Generative Conversation Skill and Fallback have been updated so that only correctly formatted URLs with https:// can be used in the "Knowledge URLs" section. Without the correct formatting, the knowledge snippets won't be updated.
Release Version v2.53 | 1 April 2024
FEATURE
We’ve added support for Romanian, Polish and Swedish so your Digital Person can speak and understand even more languages.
UPDATES
We’ve added more tool tips to the Generative Conversation Skill & Fallback to help users better understand what fields inform different aspects of the experience. The Generative Conversation Skill can also now support larger amounts of data (more than 4096 bytes) when “streaming” is toggled on.
Skills
We've been working hard to make sure the skills in our platform align with customer needs and the shift towards Generative AI based conversations. With that in mind, the following skills are in the process of being deprecated. If you use any of these skills and have questions, please email support@soulmachines.com.
Weather (Wolfram Alpha)
Twitter
Site Search
Google Form Survey
Spelling
Shopify
YouTube Channel
Greeter Base
Elegant Failure
Knowledge Base Fallback
Repeat
Facebook
Knowledge Base (Base corpus)
Joke
Contact
Translation (Microsoft Cognitive)
Math
Auto-Join Microsoft Teams
BUG FIXES
We’ve reduced the delay time when unmuting/muting a Digital Person with projects using Google’s STT.
A bug was fixed in projects using cinematic camera cuts so that URLs are pronounced correctly when using the @pronounce function.
We’ve fixed a bug so that prices with punctuation, specifically commas, are maintained as a single entity, where previously they were displayed and read as separate entities with incorrect spacing. Now, you can have your Digital Person discuss pricing with no issues!
Release Version 2.52 | 18 March 2024
FEATURE
We’ve added an improved response time by toggling on or off streaming in the Generative Conversation Base and Fallback. This allows long answers to be broken up into multiple sentences that are delivered at one time to keep the conversation engaging. By default, this button is toggled on. You can toggle it off at any point under the Generative Conversation Base settings, beneath the “Max response length” field.
UPDATES
Our Skills UI just got a new look!
Streamlined Experience: Check out the separated “Conversation” and “Additional Skills” sections under “Knowledge” when creating or editing a project - they’re much easier to navigate!
Conversation Simplified: All projects must have a “Conversation” selected, but this can be updated easily based on your project’s needs.
Easy Selection: When choosing a conversation provider, you'll find them neatly categorized as "Featured" and "External Platform" for effortless selection.
Using Orchestration? Simply select the "Orchestration Skill" in the Conversation section.
Focused Skill Selection: When adding “Additional Skills”, only non-conversation skills will appear to make it even easier to enhance your project.
BUG FIXES
We’ve fixed a bug with the Digital Chief Experience Officer (DCXO) template so the Digital Person now responds with an error-handling prompt for the user to repeat themselves.
A bug was identified and fixed in the orchestration server when no NLP was selected.
Release Version 2.51 | 4 March 2024
FEATURE
We’ve added a BETA stock Digital Person named Kate who will be wearing glasses. Kate is not editable in the Comprehensive Blender Tool just yet, but we’re working towards the ability to accessorize Digital People, and this is a great first step! Check her out in your studio account.
UPDATES
We’ve removed the Knowledge Base Builder from the main menu as this product is being moved to end of life status. If you have an existing project in Knowledge Base Builder, you can still access the URL directly.
Legacy Digital People using Human OS 2.2 or lower will be removed from studio as they’re no longer supported.
Content Cards can now be used with the Solution Expert Template as well as Generative Conversation so you can add multi-modal content to your experience.
BUG FIXES
A bug with creating video assets has been fixed so video controls will remain visible and the entire script will be recorded.
Release Version 2.50 | 19 February 2024
UPDATES
We’ve updated the Content Card Skill so that it now supports markdown tables which can be used with the Generative Conversation Base.
Release Version 2.49 | 22 January 2024
FEATURE
We’ve added support for Bahasa Malay (Malaysian) for both STT & TTS on azure speech services.
We’ve added Automatic Iconic Gestures. You can enable this for your project to take advantage of automatic gesture insertion for any conversation backend. You toggle this feature on and off under “Other Behavioral Settings” on the “Create a Project Page.” Automatic Iconic Gestures enable your Digital Person to perform additional, more expressive gestures when speaking. If these gestures don’t work in your experience, leave this toggle off. Learn more about this in Behavior Adjustments and Settings.
UPDATES
Generative Conversation Base & Fallback skill
We’re constantly updating our Generative Conversation Documentation to include the latest features and best practices for set-up.
BUG FIXES
Resolved an issue to allow URLs to show as specified in the knowledge snippets when using the Generative Conversation Base & Fallback skill
We’ve fixed an error when using the Generative Conversation Base & Fallback skill where URLs with large amounts of text were not being scraped. All URLs will now scrape up to the first 40,000 tokens. You can turn this on or off by toggling “Convert URL data into optimized conversation content” under the “Generative Conversation Base” settings.
We’ve resolved a bug with Microsoft STT timing out so connection remains intact.
We’ve fixed a bug so Digital People viewed in preview mode are now connecting correctly to OpenAI.
Release Version 2.48 | 11 December 2023
FEATURE
UPDATES
Our monthly package subscribers can now use the “Manage Subscription” button in Digital DNA Studio to update their package type.
When configuring the Content Cards Skill, you now have the option to ignore markdown syntax on image and video links - this is on by default. The Content Cards Skill can be found under “Knowledge” by clicking “Open Knowledge Manager” and then “Add Skills.”
Generative Conversation Base & Fallback skill
We’ve updated our Generative Conversation Skills Best Practices guide to include the latest tips for setting up the Generative Conversation Base & Fallback skill.
BUG FIXES
Resolved a bug where Microsoft STT failed but didn’t reconnect - will now reconnect automatically after 10 seconds.
Knowledge snippets are now being modified and summarized by OpenAI when the UI option is selected.
Release Version 2.46 | 13 November 2023
FEATURE
The Content Card Skill automatically displays contextual video, image and/or link cards to support and enhance your Digital Person's conversation responses without needing to add Soul Machines Context Variables to every conversation node as long as the conversation contains a valid public video, image & url data. This is especially helpful for generative conversations where structured context variable declaration was not possible. Simply add this skill as a “post process” skill to your favorite base conversation to supplement conversational data with multi modal content.
New Stock Digital People
New Preset Avatars available! We've refreshed our line up of Preset Avatars to show off all the cool things you can create in Digital DNA Blender. With 20 new preconfigured looks available, we're sure you'll find something that inspires!
New male tailored outfit
We have an all new tailored outfit for our masculine body types within the DDNA Blender. The outfit can be customized with preset selections or custom configurations for the inside T-shirt as well as the tailored jacket by the users!
UPDATES
Flexible Knowledge Snippet and Additional Prompt text input area in the Generative Conversation Base & Fallback skill
The Generative conversation skill iteration and editing just became a lot easier with the flexible text input area within the skill configuration modal. The area allows for natural sizing flexibility to support low or high knowledge snippet and additional prompt input to provide context to your generative conversation!
Changed GPT-TYPE configuration in the Generative Conversation Base & Fallback skill
The GPT-type drop down selection now allows users to select between “chat” & “completion” GPT-types reducing the confusion between the type of model vs model version. This also means if you are using a “Model Override”, please make sure your model override is compatible with the GPT-type (eg: if using “GPT-3.5 turbo instruct as the model override, make sure “completion” is selected as the corresponding GPT-type). Read more about it here.
No need to add “\n” to Knowledge Snippets & Knowledge URL’s in the Generative Conversation Base & Fallback skill
There is no need to separate knowledge snippets & knowledge URL’s with the “\n” new line marker as they are now automatically separated when entering knowledge snippets in separate lines within the text and URL input areas in the skill modal.
BUG FIXES
Resolved a bug where the a DDNA Studio Owner was not able to resend invitation to DDNA Studio to fellow team members who want to join their DDNA Studio organization.
Release Version 2.45 | 30 October 2023
FEATURE
The Explorer Template helps Studio Users with creating their first project. It is an “in-app helper” who can guide you through the process. It is not intended to be deployed into a production version but “uses the product to sell the product”. The Explorer is equipped with a very large store of knowledge related to Digital DNA Studio which is hosted at an external s3 link. It uses the Generative Conversation Base Skill and OpenAI’s version of ChatGPT (3.5).
Interactive Preview
A user can choose “preview” on the template card found on their Digital DNA Studio home page and begin speaking with the Digital Person in a live preview pop-up window. Out-of-the-box, the template will help users choose which smartphone is best for them based on their requirements.
STT & TTS support for Hebrew
We now offer support for our STT & TTS service in Hebrew language.
BUG FIXES
Resolved a bug where the a user with role “analytics” was not able to access Project Insights from the Project tiles on DDNA Studio dashboard
Release Version 2.44 | 2 October 2023
FEATURE
The Solution Expert Template is now available in studio. Templates serve as preconfigured Digital Person experiences, offering examples of effective Digital Person placement for success, aiming to inspire DDNA Studio users with possibilities and expedite their understanding of integrating a Digital Person into their business, service, or product. The Solution Expert Template helps users during the consideration stage of their journey and is best for situations where the decision is complex. All projects come pre-configured with the Generative Conversation Base Skill and OpenAI’s version of GPT-4 enabling the Digital Person to speak.
Preview pop-up window
A user can choose “preview” on the template card found on their Digital DNA Studio home page and begin speaking with the Digital Person in a live preview pop-up window. Out-of-the-box, the template will help users choose which smartphone is best for them based on their requirements.
UPDATE
Improved Knowledge Extraction From Customer Documentation And Websites
You can now point the Generative Conversation Skill or Base to a pdf or endpoint and it automatically creates knowledge snippets, or bits of knowledge that will be provided as context to the OpenAI GPT prompt. This allows the generative conversation skill to effectively extract knowledge from any document and website, so that you can easily train the LLM with unique content.
Skills
Skills are now available on multi-region production servers, decreasing latency and increasing resilience.
BUG FIXES
We’ve improved skills uptime and resilience
Release Version 2.43 | September 2023
FEATURE
Microsoft Power Virtual Agent (PVA)
Microsoft PVA is now supported as an NLP offering within Digital DNA Studio under “Manage Conversations and Skills.”
Skills
Skills are now available on Orchestration Servers so you can access all features when building a customized experience.
UPDATED
No Camera Idle Gazing
When users do not allow the user of their camera, the Digital Person will still perform idle gazing behavior which consists of subtle random gazing around the screen. Ideal gazing is triggered a few seconds after the user and Digital Person go quiet. If the user speaks, the Digital Person will gaze back at the user again. Behavior styles “Shy” and “Frosty” will exhibit the most visible idle gazing behavior.
Release Version 2.41 | 07 Aug 2023
UPDATE
Nine new Gesture Animations
We are excited to introduce the new insertable gestures in this release, allowing users to enhance their interactions with gestures denoted by the "#" symbol. These gestures bring a more immersive and expressive experience to the product.
Here are the available insertable gestures:
#HeartSignFingers- This gesture specifically focuses on the hand and finger movements associated with a heart sign. Use it to convey love and endearment.#Peace- With this gesture, both hands form the peace sign. It symbolizes harmony and unity, perfect for conveying peaceful intentions.#CircleContinuing- This gesture represents a continuous circular motion. Use it to emphasize ongoing or continuous actions.#ChopContinuing- This gesture involves a chopping motion, indicating a continuing action or process. It adds a dynamic element to your interactions.#Tada- This gesture showcases a celebratory "Tada" pose, expressing joy and excitement. It's great for announcing achievements or surprises.#HoldFloorHand- This gesture demonstrates a hand motion indicating holding or grasping something. It adds a sense of stability and control to your interactions.#IdeaFingerUp- This gesture features a finger pointing upward, representing an idea or a eureka moment. Use it to convey creative thoughts or suggestions.#Enthusiastic- This gesture is marked by a lively and animated demeanor, reflecting genuine excitement and passion. It involves a bright smile with enthusiastic gestures, aiming to convey and spread positive energy.#Greeting- This gesture involves friendly expressions used to acknowledge and welcome others in a cordial manner.
Updates to DDNA Designer
A new skin and makeup (casual) option has been added to DDNA Designer.
Release Version 2.40 | 10 July 2023
FEATURE
Microsoft Speech-to-Text (STT) support:
You can now choose between Google (default) or Microsoft Cognitive Services as your Speech-to-text provider. Each speech recognition provider offers a different set of languages, accents, and additional speech features. Please refer to the Language and Voice Settings documentation for more information.
UPDATE
Enhanced Listening Behaviour: Introduces improved visibility and expressiveness to existing listening behavior, including adjusted voice activity detector to reduce nodding during noise and music, and incorporates facial expressions and head motions for added acknowledgement.
Generative Conversation Skill: You can now specify a model overrides for GPT4 and ChatGPT connections, and we’ve enhanced our guardrails around user-entered configurations to prevent unusable models. Additionally we’ve improved how the Skill handles prior conversation turns - we now include other conversational turns from base conversation or other skills (when GPT skill is used as a fallback)
New Voices: We’ve now updated the available voices to include the newer voices from our supported TTS providers (Amazon, Microsoft, Google). We’ve added over 400 new voices across the supported languages for you to check out!
BUGFIX
We fixed an issue where it would appear like the Digital Person was still listening even when the microphone was muted
We fixed an issue where the Digital Person’s fingers were flickering
We fixed an issue where the Digital Person would sometimes look in the wrong direction when using Insertable Gestures
Several other small bug-fixes
Release Version 2.38 | 29 May 2023
FEATURE
New features in Digital DNA Blender
Now you have the ability to upload a background image for your Digital People. This feature allows you to further customise the environment in which your digital creations exist, providing a more immersive and personalised experience.
We have added additional phenotypes to Digital DNA Blender. This update broadens the range of characteristics available, enabling you to create even more diverse and unique Digital People.
New clothing options with increased wardrobe texture and pattern selection. For feminine Digital People, you can now dress them in a tailored jacket and for masculine Digital People, we have added more pre-curated options for the cardigan, perfect for a casual yet refined look.
Widget Source Code now available on Github
We’ve made the source code for our Widget available on Github. This means you can explore and utilize the code for your own purposes.
UPDATE
Generative Conversation (GPT) skill v1.7
The Generative Conversation skill has been released with the following new/enhanced features:
Added support for Azure OpenAI API endpoints
Changed the webscraper to only scrape the first 10 URLs presented to address an issue with excessive number of knowledge URLs overloading the API
Added ability to edit knowledge snippets after they were generated
Added flag to disable rebuild of knowledge on deploy (so any manual work doesn't get overwritten)
Up to 2 pre-built prompts can now be added to a corpus
Handle numerous edge cases that would cause the skill to not respond
Knowledge Base Builder v1.6
Knowledge Base responses now supports command and gesture markup.
Support for "Extensions" within New Watson Assistant
We’ve fixed an issue where only the second request to a Watson Assistant Extension was processed. Extensions should now work as expected. Please reach out to support if you continue to experience issues.
Release Version 2.37 | 15 May 2023
FEATURE
Multiple Organization Access
Digital DNA Studio now allows users to access multiple organizational accounts using a single set of login credentials. This feature is particularly beneficial for partner organizations that handle multiple customers and want their support users to work across different organizations.
Release Version 2.36 | 01 May 2023
FEATURE
Create DP Video Assets
We have introduced a powerful new feature: Create Video Assets. With this option, available on the DDNA Project Management page, you can generate and download videos of your Digital Person delivering a custom script. Now, you can bring your Digital Person to life in exciting new ways, and share your creations on social media to engage with your audience like never before.
UPDATE
Knowledge Base Builder
The Knowledge Base builder has been enhanced with the ability to save the knowledge base in the application itself allowing you to come back to it and continue work.
Knowledge Bases are now stored in the cloud, enabling access and use within DDNA Studio without updating Knowledge Base Skill configuration manually for every change to the corpus.
We’ve added the functionality to ensure knowledge bases produced via Knowledge Base Builder will populate the Insights Dashboard out of the box, no action required.
New clothing options for Digital People
We now have introduced blazer and boatneck outfit for feminine Digital People.
Release Version 2.35 | 03 Apr 2023
FEATURE
Generative Content Skill v1.2
We are thrilled to announce the latest release of Generative Conversation Skill v1.2! This highly anticipated update includes support for ChatGPT and GPT-4, enhancing the already impressive capabilities of Generative Conversation Skill.
v15.7 Soul Machines WebSDK
FEATURE
WebSDK now supports the ability to not stop DP from speaking when switching tabs. A new optional parameter stopSpeakingWhenNotVisible has been added to the Scene Configuration. When true, the Digital Person's speech will be interrupted if the user changes tabs or navigates to another page. This will prevent the Digital Person from reaching any content cards or speech markers that have not yet been reached at the time of interruption. Set to false to allow DP to continue speaking while a different tab is active.
Release Version 2.34 | 21 Mar 2023
FEATURE
Generative Content Skill v1.1
Well, Generative AI has certainly been in the news lately, so we're excited to announce our Generative Conversation Skill. This Skill lets you connect your Digital Person to the GPT-3 Large Language Model (LLM) and provide knowledge snippets for use by the model. This is the same Skill that powers Suki on our website.
New clothing options for Digital People
We now have introduced full sleeve formal shirts, scrubs, and sweaters with a wide variety of patterns and colours for clothing.
Single Sign-on login for Soul Machines Ecosystem
DDNA users can now navigate to SM Academy, Creator Community, Support, and KnowledgeBase Conversation Builder using the Soul Machines credentials. To support this feature, the account name has been split into First and Last names on the Studio Account page.
Create and Manage Skills from Digital DNA Studio
You can now create and manage skills for users within your organization. The purpose of creating a skill is to register it on the Soul Machines Platform. Any user with an owner, support, or editor role, can create skills, and anyone in the organization can edit Skill definitions regardless of whether they created them.
UPDATE
Voiceflow v1.2
This release now includes support for the Speech Modality, joining our existing support for Text modality.
Skin Swatches in DDNA Blender
We have updated the skin tone swatches in DDNA Blender to give more contextual information about each skin option, including whether it includes facial hair and the eyebrow shapes.
BUG FIX
Survey Skill v1.3
This is primarily a bug-fix release that improves matching of choice questions when the answer does not exactly match the option from the original form. It also improves usage as a Base conversation with greeting support.
Release Version 2.33 | 20 Feb 2023
UPDATE
KnowledgeBase Builder 1.4
The knowledge base builder has been enhanced to allow training and testing of conversations, making it easier and faster to iterate.
Support for Voiceflow Content cards
We now support visual features of voiceflow through content cards. This does require that you use a VoiceFlow "Custom Action" step, which is only available if you have a voiceflow subscription (not available for free accounts).
Release Version 2.32 | 07 Feb 2023
FEATURE
Voiceflow Support 1.0
Conversation Designers rejoice! We now support Voiceflow as a conversation provider. VoiceFlow provides an easy to use visual editor for creating conversations. V1.0 offers basic support for text conversations. Future releases will add further features.
UPDATE
KnowledgeBase Builder 1.3
We've improved the user experience when editing the knowledge base. Please keep your feedback coming!
BUG FIX
Digital Person Under Eye Wrinkle
We’ve resolved a bug impacting an over exaggerated wrinkle on some Digital People created with Comprehensive DDNA Blender.
Release Version 2.31 | 23 Jan 2023
FEATURE
Project Preview
Users of DDNA Studio can now preview their Digital Person configurations all within DDNA Studio Project configuration sections! The project configuration area will showcase the preview on the right half of the screen with configuration options and selections to the right of the preview. After making selections such as a look, language, voice, behavior etc. users can click “update preview” at any time to see their changes immediately and choose appropriate configurations for their final experience. In addition, the Preview also offers a Script for the users to type text to use as part of Digital Person narration within the Preview.
v1.12 Web Widget
FEATURE
Web Widget now supports full frame layout and float layout which can be toggled by conversation by the end user to draw attention to the Digital Person on the widget.
Expose scene and persona as public properties of the widget
BUG FIXES
Fixed audio button color for muted state
Fixed firefox scroll issue when going from a long content card to a short content card
Fixed general issues with markup card
Fixed the outline of the content card
Fixed Safari bfcache issue - reload when navigating from bfcache
v15.6 Soul Machines WebSDK
FEATURE
WebSDK now has tracking capability to measure Digital People connection time tracing
We now have support for logging API key errors with hints
BUG FIXES
ContentAwareness change unable to find video error to warning
A bug causing issues with ContentAwareness support within session-persistence is now been fixed
Release Version 2.30 | 14 Dec 2022
FEATURE
Knowledge Base Builder
We’ve now released the first version of our Knowledge base builder - a GUI that allows creators to easily create a knowledge base without needing to edit JSON, but instead via an intuitive user interface. Included in this release is the ability to Crawl a website and pre-populate the KB JSON file with all the pages found, along with the meta:description / H1 / title of the page as a starter response. The intention here is greet the end-user with a page-specific response as KnowledgeBase will match on the page URL as they navigate the site. The Knowledge Base Builder can be accessed at https://knowledgebase-ui.dev.soulmachines.cloud/.
Shopify Skill
With the Shopify Skill, you can interact with a digital person while browsing a Shopify website. With this skill, users can search for products and then navigate to the page of the item. The user can ask questions about a product, such as: “What sizes does this come in?”, “Is black available?”. The user can ask to see their cart, and the skill will navigate to the user’s cart. If the user is on a product page, they can request it to be added to their cart.
Release Version 2.29 | 28 Nov 2022
FEATURE
DDNA Blender: Significant Update: Next-Gen CGl Platform
We are invested in making world-class CGI and have updated our CGI platform to deliver:
Future-proofing for the Metaverse
Significantly increased realism of skin detail
More “true to life” skin colors
Better handling of light and shadows
Rollout details:
The new platform is opt-in. Re-publish your Digital Person from DDNA Blender to take advantage of this new platform
DDNA Blender: Skin Luminosity Sliders
We’ve added the ability to modify the brightness and the gloss/matte levels for the skin texture
Thai language support
Our Digital People can now understand and speak the Thai language.
Joke Skill
The Joke skill when enabled and triggered will randomly draw a joke from a collection of jokes.
The skill when enabled and triggered will randomly draw a joke from a collection of jokes. Using the Skill Configuration, you can toggle on and off the "Only return clean jokes" setting, which filters what jokes can be returned.
Soul Machines MobileSDK
We have now made the following sample project GitHub repos public to provide a starting point for developing custom mobile UI.
Kotlin sample for Android: GitHub - soulmachines/smandroidkotlin: Soul Machine Android SDK Kotlin Sample App
Java sample for Android: (smandroidjava) Soul Machines Android Java Sample Project
iOS sample app: GitHub - soulmachines/smiossamples: Repository for iOS Samples
Release Version 2.28 | 14 Nov 2022
FEATURE
Mobile API keys
DDNA Studio API Keys page now allows you to create Mobile API keys. The Mobile API Keys provide easy authentication to connect any Soul Machines APIs and Mobile devices.
UPDATE
Knowledge Base Skill
The KnowledgeBase Skill now supports session persistence on a website when used in conjunction with the Web Widget. This allows you to configure per-page (or wildcard) responses that will be spoken by the Digital Person when the user navigates to the page with an active Digital Person.
NOTE: This will only work with sm-web-component based sites or custom front-ends that send the PAGE_METADATA message.
NLP Adapter Skill Documentation
We’ve updated the getting started documentation for NLP Adapter Skills to guide you through integrating Soul Machines Skill API with a third-party NLP system or chatbot platform.
v1.3 Soul Machines MobileSDK
FEATURE
API Key support and OS Updates
The Android and iOS SDKs have both been updated to v1.3, which includes the following features:
Android SDK support for Android 13
iOS SDK support for iOS 16
API Key support for both platforms
v1.11 Web Widget
BUG FIXES
Widget bundling issue with global variable conflicts has now been resolved
Video autoplay issue on the Safari browser has now been resolved
For certain deployments, the default profile background color was transparent, this issue is now resolved
Release Version 2.27 | 31 Oct 2022
UPDATE
Survey Skill
The Google Forms Survey Skills now supports the following features of Google Forms:
Checkboxes
Title and description
Sections
Images as options
Accepts multiple answers for checkbox question type
Note: To display images as options and support multi-select via mouse in the UI, you should either implement a custom UI or contact your customer representative to request a beta version of the web component which does have it enabled.
Release Version 2.26 | 17 Oct 2022
FEATURE
Conversation Visualization
DDNA Studio users can now visualize the path each conversation takes through the Conversation Node Graph. The node graph feature, available under the Insights tab of each project provides visibility to the different paths for users in the conversation along with insights on the most frequently visited nodes.
Visibility of HumanOS version
The avatar tiles within the Project configuration area of DDNA Studio now indicate the HumanOS version of the Digital Person. This tag helps you understand the features/configurations available for a Digital Person depending on the HumanOS version with which they are built. The indicator also helps users to visibly know if the Digital Person is on the latest version of HumanOS.
Project Configuration Options Notification
The project configurations area now features improved sidebar indication of form errors and inline messages for unavailable sections of the configuration to give DDNA Studio users details about available and unavailable sections along with tooltips to enable the use of those sections.
As an example, Studio now shows an inline message in the Behavior Section of a project having a Digital Person with an older version of HumanOS to let the user know that Behavior Styles section is not available for this HumanOS and similarly a red error indication appears in the sidebar if a section of the configuration has an error associated with it.
UPDATE
Survey Skill
The Survey Skill has received several updates including:
The ability to implement conditional question logic using a webhook
The ability to turn a ‘survey’ into a ‘quiz’ using a webhook
Web API Keys are now editable
The web API keys can now be edited post-creation, allowing for configuration changes like editing the applicable domains and projects. This can now be done by navigating to the API keys section of Studio and clicking the edit icon.
v1.10 Web Widget
FEATURE
The default logs for the web widget can now be turned off, additionally, live logging configuration can be enabled
The advanced CSS customizations can now be easily previewed, configured, and copied through a live preview of the CSS configurations on a web widget. Try this feature here.
v15.3 Soul Machines WebSDK
FEATURE
WebSDK now supports the ability to turn off default connection logs and enable developers to allow live logging configuration for scene and content awareness from the Scene constructor. The available log levels include
debug,warn,log, anderrorwith the ability to define the minimum log level and change it live during the scene.
Release Version 2.25 | 03 Oct 2022
FEATURE
Microsoft Teams Auto Join Skill
This Skill, when configured with a Digital Person, will join a Microsoft Teams meeting when invited. The skill will monitor the DP’s Microsoft Outlook schedule and if a meeting is approaching (roughly 2-3 minutes away) the Skill will have the DP join the meeting. This Skill will check for a meeting every 60 seconds. The skill will constantly check their calendar until the Digital Person is deleted from DDNA studio.
UPDATE
Site Search Skill
Now can return up to 4 internal links card types, which enables the Digital Person to surface cards that allow the user to navigate to other pages on the same domain.
Knowledge Base Skill
The knowledge base skill now has a new extension that allows for prebuilt models to be included. Our first prebuilt model is Elegant Failure, which mitigates user frustration by intelligently responding to any request your conversation can’t answer.
Release Version 2.24 | 19 Sept 2022
FEATURE
New Digital CXO Role
Our Digital CXO is intended to conduct 1:1 customer research at scale. Most useful for CSAT or NPS type scores, a DCXO can help you understand what your website visitors are really thinking beyond what can be captured in traditional methods (eg, web forms). Powered by a simple integration with Google Forms, you can get a Digital CXO up and running on your website in less than an hour.
Contact Skill
Our Contact Skill allows you to now collect information from the user for follow-up. We currently support Name, Phone number, and Email address.
New Weather Skill
We’ve added a new Weather Skill which uses OpenWeatherMap, which has images for different weather conditions.
Studio Notification System
DDNA Studio now offers in-app notifications that allow Soul Machines Admins to notify users of critical or relevant information, warnings, and errors. The notifications appear at the top right bar of Studio Dashboard through the bell icon.
DDNA Studio Invitation
Studio organization owners and support user roles can now resend DDNA Studio invitations to other members in their organization if the original link to the invitation has expired.
UPDATE
Wake Word Skill
We now support a comma-separated list of wake words & sleep words, as well as allow you to customize what the DP says when it is asked to go to sleep. We’ve also now added Initial and Default states.
Survey Skill
The Survey Skill now supports showing an image per question (note, the image per answer feature is coming soon)
Insights API
UPDATE
Integration guides
Insights API documentation has been updated to include a step-by-step guide for connecting Insights API as a data source to DOMO and PowerBI for easy integration.
v1.9 Web Widget
FEATURE
Web Widget now displays a loading bar indicator showing the progress of the connection status to your Digital Person.
UPDATED
Widget now supports a simplified schema for advanced CSS customizations. The palette now supports grey, primary, secondary, and tertiary color customizations.
v15.2 Soul Machines WebSDK
FEATURE
We now have a new Connection State Event API that communicates loading progress in five different states of connection as
ConnectionStateTypes. These events can be useful for developers building branded and custom-loading experiences using the webSDK.
Release Version 2.23 | 05 Sept 2022
FEATURE
New NLP Platform
You can now connect New Watson Assistant projects to your Digital Person
Improve Navigation on the Project Form
Paginate and divide the project form into sections for quick project configurations by letting the user directly jump to the corresponding section of the form
UPDATE
Knowledge Base Skill
Now includes support for content cards using the standard
showCard()syntax. Read more here.Now includes a Call To Action response which the DP can speak after replying to the user’s query. For example: “what else can I help you with?”
BUG FIX
We’ve done some work behind the scenes to minimize issues where the Digital Person’s body might show through their clothing
Increase Session Server health check interval to avoid blender being killed on startup
v15.1 Soul Machines WebSDK
UPDATE
Support for Single Page Applications with SM Digital Person to send
PAGE_METADATAintent to NLP engine whenever the URL changes
Release Version 2.22 | 22 Aug 2022
FEATURE
Role Templates in DDNA Studio
Introducing the Digital Workforce
Now getting setup with a Digital Employee is even easier with our new Role templates, which allow you to have a fully functional Digital Person running on your website with just a couple of clicks.
Our first role is the Digital Greeter. Try it out in DDNA Studio today! The templates are available to select on the DDNA Studio Dashboard page above your projects section. The templates can be customized or utilized as it is.
UPDATE
Event Scheduling Skill
The Event scheduling skill now displays a markdown graphic of the user’s calendar with the event. Users can also specify phrases such as "today", "tomorrow", "noon", “next monday”, or certain US holidays to schedule events.
DDNA Blender
We’ve added 12 new hairstyles, including making all existing hairstyles come in at least 7 colors, significantly increasing the variety of looks possible
BUG FIX
Fix for some hair fitting incorrectly in DDNA Blender
Resolved an issue with the Repeat skill where the Digital Person repeated what the user said rather than their last response, as well as the content cards not displaying properly.
v1.8 Web Widget
FEATURE
Web Widget now has a new Turn Taking indicator within its UI to visualize the state of speech and thinking behavior by highlighting when user speech is recognized, when the digital person is speaking, and then when the digital person is thinking or processing the user’s input.
We now have a “bottom left” positioning option for the widget using the parameter
data-sm-positionand setting it tobottomLeft. This feature expands the options to position the widget between the bottom right and bottom left.The version of the web widget can now be easily known through the
srcparameter within the embed code
BUG FIXES
Fixed an issue with Internal Link content cards reappearing
v14.6 Soul Machines WebSDK
FEATURE
Support for new API for listening events as
ConversationStateTypesto indicate four different conversation states: “dpprocessing”, “dpspeaking”, “userspeaking” and “idle”.
UPDATE
“PAGE_LOADED” intent is now updated to “PAGE_METADATA”
“PAGE_METADATA” intent is not sent to the conversation at the conversation start to avoid any potential confusion with the “Welcome” intent
Release Version 2.21 | 08 Aug 2022
FEATURE
Digital Greeter Base Conversation
This Skill is the Base for the Greeter Role. It introduces the Digital Person, welcomes the user with optional rapport building, and sets framing for the conversation.
Site Navigation Skill
The Site Navigation Skill enables the Digital Person to search a domain name for particular content, if no matches are found in other Skills. The top result is shown to the user as a link card so the user can navigate to that content.
UPDATE
Stock DPs upgraded to HumanOS 2.4
The “Stock” Digital People are now running HumanOS2.4, so you can take advantage of all the latest features
Updates to Skills
The Google Forms Survey Skill now allows you to order the linear scale question in ascending or descending order.
BUG FIX
Removed Misleading "Number of Conversations"
The number of conversation counts shown under the “Accounts and Team“ page have been removed in lieu of the number showing incorrect values
v1.7 Web Widget
FEATURE
Web Widget now supports Markdown Card as a valid content card. The markdown card, once configured can be used to present content in plaint text formatting (markdown) that gets converted to html content. A common use of the markdown card is to present a variety of different types of information together in one card and customize the card’s content
Web Widget users can now use Advanced customization configuration to customize the default colors, styles and fonts by using css variables specific to the widget in their website’s style sheet. The customizations currently possible include Theme, Font, Color Palette and Box Shadow.
v14.5 Soul Machines WebSDK
FEATURE
Support new intent “PAGE_LOADED” and context variable “pageUrl“ to the NLP denoting the status of a new page load and the page url respectively. The “PAGE_LOADED” intent is sent upon the detection of a new page load in an active Digital Person session. The “pageUrl” context variable captures the URL of the current webpage. The context variable updates for each “PAGE_LOADED” intent.
Release Version 2.20 | 25 July 2022
FEATURE
Digital People
You can now collect the DNA from any existing Digital Person within your organisation from the 3-dot menu on the project form page. This feature works for any Digital People created via the comprehensive Blender since its release. The DNA for Digital People created via the Simple mode can be found on the DNA Bank page.
Microsoft Events Skill
This Skill gives the Digital Person the ability to conversationally schedule a Microsoft Outlook meeting. The Skill is triggered by Create/Setup/Schedule (an event)/(a meeting).
You can ask for a specific person (firstname, lastname) to be added to the meeting (name must be an active user in your organisation or in your contacts). By default is 30 mins meeting is created at the time requested. The Digital Person will request for the meeting title. Currently limited to entering only one person or one group at a time. End user is asked if they want to add anyone else, saying “no” ends the Skill.
Knowledge Base Skill
This Skill gives the Digital Person the ability to answer end user questions based on a customer’s knowledge base. The Knowledge Base Skill data needs to be in a JSON file that contains three fields for each record:
“intent” - this can be anything as long as it is unique for the project
“examples” - This is an array of strings that contain examples of different phrasings for the question
“metadata” - This contains an array of strings that are the responses to the question. If more than one value is found for a given intent, the skill will randomly pick one of the responses to reply with.
If there is not an exact match, a NLU process will attempt to find the closest intent based on the examples. If no match is found, it should reply back with a configurable NO_MATCH response.
UPDATED
Skills
Additional configuration options have been added to the Google Forms Survey Skill: you can now disable Digital Person from reading all options of multiple choice questions and display upper and lower bounds of linear scale questions.
Support for Default Values in Skill Configuration
BUGFIX
Fix for Skill_Config not available in Watson first conversation turn
v1.6 Web Widget
FEATURE
Web Widget now supports Internal Link Card as a valid content card. The internal links card, once configured appears on top of the Digital Person Widget as a floating card to present a clickable button associated with a link and optionally an image and description to go along with the link. The button when clicked loads the page in the same tab as the current webpage
UPDATED
The icon to represent a disconnected state has been replaced by ‘X’ for a clearer representation
The image shown in the image card is now in full size
BUG FIXES
Fixed widget video remaining black on iPadOS/iOS Safari 15.5
Fixed incorrect color and states of camera and microphone
Release Version 2.19 | 11 July 2022
FEATURE
New Skill: Transcription
When a DP has the Transcription Skill, the conversation session is stored as a text file in the Customer’s secure Azure S3 container with the user's requests and conversation responses. The transcript is update throughout the conversation.
New Skill: Wake Word
This Skill was designed to help the Digital Person when there are multiple end users talking. It puts the Digital Person into a Sleep Mode, until it hears the predefined Wake Word, and vice versa. This prevents the DP from interrupting in meetings where multiple people are speaking.
Example:
WAKE WORD: Hey Alice
READY WORD: Alice is online. How can I help.
SLEEP WORD: Good bye Alice
New Skill: Repeat
The Repeat Skill simply repeats the very last response from the Digital Person.
DDNA Blender
Created a rad looking Digital Person and want to share it? Or perhaps after you published it you wanted to edit it later, but couldn’t? Now both of these things are possible with our new “DNA” feature which allows you to encapsulate all the Blender settings that created that specific look. You can send DNA to other DDNA Blender users so they can recreate the same look or use it as a starting point for iterations.
In this release we are launching the ability to use DNA from within the Blender. A future release will extend this to the project page as well.
Web API Keys
Session persistence and related cross page authentication for pages within the same domain is now supported by web API keys through DDNA Studio. This feature requires regenerating the web API keys for respective projects.
UPDATED
Insights API
The “Conversation Node Summary” endpoint now supports a new “graph” mode which makes it easy to support popular graphing libraries like Apache E-Charts or D3.js. This allows:
Data filtering set that can be requested
Pre-process the data for the “graph response” of the endpoint to avoid circular references due to conversation loops, for Sankey diagrams.
A special conversation start and end node, for cleaner graphs.
v1.5 Web Widget
FEATURE
The below features for the Web Widget are not backwards compatible and require a re-embedding of the copy pasteable embed snippet from DDNA Studio
Web Widget now supports External Link Card as a valid content card. The card follows the same command as the Default UI (microsite) deployment. The external links card, once configured appears on top of the Digital Person Widget as a floating card to present a clickable button associated with a link and optionally an image and description to go along with the link.
Web Widget now supports Image Card as a valid content card. The card follows the same command as the Default UI (microsite) deployment. The image card, once configured appears on top of the Digital Person Widget as a floating card to present a remotely hosted image
Web Widget now supports Session Persistence through Web API Key authentication. This feature allows the widget to persist and keep the conversation continuous between page loads of the same domain within the same tab of the browser.
KNOWN ISSUES
Widget video remains black on iPadOS/iOS Safari 15.5 upon entry. This is resolved by clicking any button within the widget UI
MacOS Monterey 12.4 Safari 15.5 - Digital Person in the web widget is muted when navigating to a different page in the same tab (using session persistence). This is Safari autoplay policy limitation, when navigating to next page, the video can only be played muted, user needs to unmute it from button.
In Safari, user is prompted for media permission after navigation if media was enabled prior to navigation. Advanced users may change this behavior using Safari → Settings for [Site Name] → Microphone (and/or camera) → Allow
v14.4 Soul Machines WebSDK
FEATURE