We have seen significant need from our customers and partners to deliver a conversational assistant tailored to their brand, personalized to their customers and made available across a broad range of conversational canvases and devices. Continuing Microsoft open-sourced approach toward Bot Framework SDK, the open source Custom Personal Assistant solution provides full control over the end user experience built on a set of foundational capabilities. Additionally, the experience can be infused with intelligence about the end-user and any device/ecosystem information for a truly integrated and intelligent experience.
We strongly believe our customers should own and enrich their customer relationships and insights. Therefore, any Virtual Assistant provides complete control of the user experience to our customers and partners through open-sourcing the code on GitHub. The name, voice and personality can be changed to suit the organization’s needs. Our Virtual Assistant solution simplifies creation of your own assistant enabling you to get started in minutes and then extended using our end to end development tooling.
The scope of Virtual Assistant functionality is broad, typically offering end users a range of capabilities. To increase developer productivity and to enable a vibrant ecosystem of reusable conversational experiences, we are providing developers initial examples of reusable conversational skills. These Skills can be added into a conversational application to light up a specific conversation experience, such as finding a point of interest, interacting with calendar, tasks, email and many other scenarios. Skills are fully customizable and consist of language models for multiple languages, dialogs and code.
At this time we are running an initial preview and working closely with initial customers and partners in an open-source repository to bring to life the first experiences and make it available more broadly in the coming months.
The Virtual Assistant builds on the Enterprise Template capabilities which provide foundational capabilities for conversational experiences including base conversational intents in multiple languages, Dispatching, QnA and conversational insights. The following Assistant related capabilities are provided at this time, further capabilities are planned and we'll be working closely with customers and partners to help inform the roadmap.
| Feature | Description |
|---|---|
| Onboarding | An example OnBoarding flow enabling your Assistant to greet the user and collect initial information. |
| Eventing Architecture | Events in the context of the Virtual Assistant enable the client application hosting the assistant (in a web-browser or on a device such as a car or speaker) to exchange information about the user or device events whilst also receiving events to perform device operations. |
| Linked Accounts | In a speech-led scenario it's not practical for a user to enter their username and password for supporting systems through voice commands. Therefore a separate companion experience provides an opportunity for the user to signin and provide permission for an Virtual Assistant to retrieve tokens for later use. |
| Skill Enablement | A broad set of common capabilities exist which today, which require each developer to build themselves. Our Virtual Assistant solution includes a new Skill capability enabling new capabilities to be plugged into an Virtual Assistant through configuration only and provide an authentication mechanism for Skills to request tokens for down-stream activities. |
| Point of Interest Skill | The preview Point of Interest (PoI) skill provides a comprehensive language model for finding points of interest and requesting directions. The skill currently provides integration into Azure Maps. |
| Calendar Skill | The preview Calendar Skill provides a comprehensive language model for common calendar related activities, The skill is currently integrated into Microsoft Graph (Office 365/Outlook.com) with support for Google APIs to follow soon.. |
| Email Skill | The preview Email Skill provides a comprehensive language model for common email related activities, The skill is currently integrated into Microsoft Graph (Office 365/Outlook.com) with support for Google APIs to follow soon. |
| ToDo Skill | The preview ToDo Skill provides a comprehensive language model for common task related activities, The skill is currently integrated into Microsoft Graph (outlookTask). |
| Automotive Skill | The preview Automotive Skill provides Vehicle Setting control support for an assistant along with an initial set of common vehicle settings which can be extended further. At this time the skill surfaces device events to demonstrate how device control could be achieved. |
| Device Integration | Our Azure Bot Service SDKs (Direct Line) along with Adaptive Card and Speech SDKs enable easy cross platform integration to devices. Additional device integration examples and platform including Edge are planned. |
| Test Harnesses | In addition to the Bot Framework Emulator, a WebChat based test harness is provided enabling more complex authentication scenarios to be tested. A simple Console based test harness demonstrates the approach to exchange messages to help frame the ease of device integration. |
| Automated Deployment | All the Azure resources required for your Assistant are automatically deployed: Bot registration, Azure App Service, LUIS, QnAMaker, Content Moderator, CosmosDB, Azure Storage, and Application Insights. Additionally, LUIS models for all skills, QnAMaker, and Dispatch models are created, trained, and published to enable immediate testing. |
| Experimental Skills | Experimental Skills are early prototypes of Skills to help bring skill concepts to life for demonstrations and proof-of-concepts along with providing different examples to get you started. Restaurant Booking and News Skills are currently available. |
| Conversational Analytics | Analytics surfacing insights into end user activity, dialog/skill usage, quality and gaps in knowledge. |
The Virtual Assistant extends across a broad number of industry scenarios, some example scenarios are shown below for reference purposes.
-
Automotive Industry: Voice enabled Personal Assistant integrated into the car providing end users the ability to perform traditional car operations (e.g. navigation, radio) along with productivity focused scenarios such as moving meetings when your running late, adding items to your task list and proactive experiences where the car can suggest tasks to complete based on events such as starting the engine, traveling home or enabling cruise control. Adaptive Cards are rendered within the Head Unit and Speech integration performed through Push-To-Talk or Wake Word interactions.
-
Hospitality: Voice enabled Personal Assistant integrated into a hotel-room device providing a broad range of Hospitality focused scenarios (e.g. extend your stay, request late checkout, room service) including concierge and the ability to find local restaurants and attractions. Optional linking to your Productivity accounts open up more personalised experiences such as suggested alarm calls, Weather warnings and learning of patterns across stays. An evolution of the current TV personalisation experienced in room today.
-
Enterprise: Voice and Text enabled branded Employee Assistant experiences integrated into enterprise devices and existing conversation canvases (e.g. Teams, WebChat, Slack) enabling employees to manage their calendars, find available meeting rooms, find people with specific skills or perform HR related operations.
All aspects of the end user experience are owned and controlled by you. This includes the Branding, Name, Voice, Personality, Responses and Avatar. The source-code to the Virtual Assistant and supporting Skills are provided in full enabling you to adjust as required.
Your Virtual Assistant will be deployed within your Azure subscription. Therefore all data generated by your assistant (questions asked, user behaviour, etc.) is entirely contained within your Azure subscription. See Cognitive Services Azure Trusted Cloud and the Azure section of the Trust Center more specifically for more information.
The Virtual Assistant leverages the Microsoft Conversational AI platform and therefore can be surfaced through any Bot Framework channel – e.g. WebChat, FaceBook Messenger, Skype, etc.
In addition, through the Direct Line channel we can embed experiences into Desktop and Mobile Apps including devices such as Cars, Speakers, Alarm Clocks, etc.
The Virtual Assistant solution is built on the Azure Bot Service, Language Understanding Cognitive Service, Unified Speech along with a broad set of supporting Azure components meaning that you benefit from the Azure global infrastructure including ISO 27018, HIPPA, PCI DSS, SOC 1,2 and 3 certification.
In addition, Language Understanding support is provided by the LUIS Cognitive Service which supports a broad set of languages listed here. The Translator Cognitive Service provides additional Machine Translation capabilities to extend the reach of your Virtual Assistant even further.
Your Virtual Assistant can be integrated into your device and ecosystem enabling a truly integrated and intelligent experience. Through this contextual awareness more intelligent experiences can be developed and deliver further personalisation than otherwise possible.
The Virtual Assistant enables you to deliver your own unique experience but also handoff to the end user's chosen Digital Assistant for certain types of questions.
Our Virtual Assistant architecture is flexible and can be integrated with existing investments you may have made into device-based Speech or Natural Language processing capabilities and of course integrate with your existing back-end systems and APIs.
Adaptive Cards provide the ability for your Virtual Assistant to return User Experience elements (e.g. Cards, Images, Buttons) alongside text base responses. If the device or conversation canvas has a screen these Adaptive Cards can be rendered across a broad range of devices and platforms providing supporting User Experience where appropriate. Samples of Adaptive Cards are available to try out, as well as documentation on how to render cards within your own application.
In addition to the base assistant, there exists a broad set of common capabilities which require each developer to build themselves. Productivity is a great example where each organisation would need to create Language Models (LUIS), Dialogs (Code), Integration (Code) and Language Generation (Responses) to enable popular Calendar, Task or email experiences.
This is then further complicated by the need to support multiple languages and results in a large amount of work required for any organisation building their own assistant.
Our Virtual Assistant solution includes a new Skill capability enabling new capabilities to be plugged into a custom-assistant through configuration only.
All aspects of each Skill (Language Model, Dialogs, Integration Code and Language Generation) are completely customisable by developers as the full source code is provided on GitHub along with the Virtual Assistant.
Initial previews of the Email, Calendar, ToDo, Point of Interest and Automotive skills are now available.
- Virtual Assistant
- Skills
This project welcomes contributions and suggestions. Most contributions require you to agree to a Contributor License Agreement (CLA) declaring that you have the right to, and actually do, grant us the rights to use your contribution. For details, visit https://cla.microsoft.com.
When you submit a pull request, a CLA-bot will automatically determine whether you need to provide a CLA and decorate the PR appropriately (e.g., label, comment). Simply follow the instructions provided by the bot. You will only need to do this once across all repos using our CLA.
This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact [email protected] with any additional questions or comments.
Copyright (c) Microsoft Corporation. All rights reserved.