Store Assistant is an AI-powered product discovery and support chat tool for visitors to Microsoft.com.
It can help you learn more about products, identify which products are likely to meet your unique needs, and help with basic support questions.
Note: The Store Assistant uses the latest Large Language Model (LLM) from Azure OpenAI to augment your site experience through natural chat interaction. Learn more about Artificial Intelligence Solutions from Microsoft AI.
Store Assistant helps users by:
-
Identifying the products that are most likely to meet the needs you describe.
-
Providing information on whether a product is in stock.
-
Locating promotions and deals.
-
Helping to make product comparisons.
-
Locating Microsoft site resources, including support information or policies.
-
Providing technical support and troubleshooting steps.
If preferred, Store Assistant also makes it easy to chat with a human Store Associate for purchase-related inquiries and can connect you to technical support.
Store Assistant is best used to help find the products and services that suit your needs. Generic questions are best handled by Microsoft Copilot.
Store Assistant is powered by Azure Open AI models, trained on natural language text and source code from publicly available sources. Store Virtual Assistant also uses Bing Custom Search Web Search and Microsoft.com's Product Search for enriching each conversation with up-to-date information.
This means that AI systems are not perfect. They can make mistakes or produce false or misleading information. For example, AI systems can confuse facts and fiction; invent details that are not based on reality; fail to capture the style, preferences, and creativity of the user, or be influenced by biases or hidden agendas of the data or the developers.
All conversations begin with speaking with the AI. Additionally, you will see a welcome message informing you that you are interacting with an AI.
Before connecting to a person from the Store Associate team, you will be informed that you will be connected. When the chat window title displays a banner indicating you are connected and the Store Associate’s name, this indicates you are interacting with one of our human Store Associates.
Store Assistant has been assessed for use in English markets for its initial release. Non-English speakers might experience a lower quality of service.
This chat experience is intended to provide sales assistance and support. If you wish to speak with a Microsoft representative, simply indicate what your needs are for (e.g., “Human Technical support for Surface hardware”) in the chat and the assistant will provide you with the available options for your stated need.
Sales assistance can be transferred to a person from the Store Associate team within the same chat window. All other inquiries such as technical support will be provided links to official Microsoft Support sites where you can obtain assistance from the available options.
Privacy
Store Assistant has been built on the Azure Open AI service which processes some data related to service use, including prompts submitted by users. For example, the Azure Open AI Service includes a content management system that works alongside the models to filter potentially harmful content. This system works by running both the input prompt and generated completion through an ensemble of classification models aimed at detecting misuse.
The Azure Open AI Service also stores prompts and completions for up to 30 days to monitor content and/or behaviors that suggest use of the service in a manner that may violate applicable product terms.
For more information, see the "Data, privacy and security" section of the Azure OpenAI documentation.
Detailed information on data practices and processing in Azure OpenAI is available in their documentation.
In addition to Azure OpenAI uses, Store Virtual Assistant collects limited telemetry data used by Microsoft to monitor service quality and enable improvements. Such uses may include:
-
Evaluating the AI chat's efficacy (e.g., by measuring usage, session length, and the number of questions asked and answered per session). Investigating and detecting potential abuse of Store Assistant.
