Responsible AI
In ALVAO SaaS, AI functions are available only in the Enterprise edition.
No AI function autonomously performs actions on behalf of a user.
The following ALVAO functions involve AI:
| AI function | Description | AI accessed data |
|---|---|---|
| Ticket - Assistant tab | Allows users to find similar tickets or possible solutions, detect major incidents, and get a summary of the communication. Included in the ALVAO Service Desk AI Assistant module. | Tickets (including ticket log records, such as communication records) contained in configured ticket services; knowledge articles. |
| Service Desk - Daily summary emails | Generates a daily summary of service ticket updates. Included in the ALVAO Service Desk AI Assistant module. | Tickets (including ticket log records, such as communication records) contained in configured ticket services. |
| Knowledge Base AI Assistant | Automatically generates knowledge base articles about recurring issues and proposes solutions. Included in the ALVAO Knowledge Base AI Assistant module. | Tickets (including ticket log records, such as communication records) contained in configured ticket services; knowledge articles. |
| Mobile app - New asset - OCR | Allows users to fill asset properties with text extracted from a camera image. Included in the ALVAO Asset Management Intelligence module. | An image taken by the mobile device camera in the application. |
| Objects - New object - From document | Allows users to create new objects by extracting information from an invoice or delivery notice in PDF format. Included in the ALVAO Asset Management Intelligence module. | An uploaded PDF document. |
| Tickets - searching through images | Enhances ticket search by searching through images that are part of or attached to ticket log records (for example, screenshots capturing error messages). Included in the ALVAO Service Desk Intelligence module. | Images that are part of or attached to ticket log records. |
| Semantic search | Enhances searching by enabling searches using synonyms. Included in the ALVAO Service Desk Intelligence module. | Tickets, objects, news posts, knowledge articles. |
| Keyword search | Improves search by searching only keywords extracted from a user-supplied query in ALVAO. Included in the ALVAO Service Desk Intelligence module. | User-supplied search query. |
| Chat intent analysis | The ALVAO for Teams bot analyzes the user's intent from the text entered and acts on it. Included in the ALVAO Service Desk Intelligence module. | User's prompt in the chat. |
| User Success AI Agent | User chat with an AI agent. See the ALVAO User Success AI Agent module. | A user's prompt in the chat; tickets, objects, news posts, and knowledge articles accessible to the current user. |
| Automatic Content Translation | Translates data in ALVAO to various languages. See the ALVAO Automatic Content Translation module. | ALVAO configuration data (service names, ticket statuses, and similar items); news posts; knowledge articles; ticket log records; ticket property values; messages composed in some ticket actions. For more details, see the ALVAO Automatic Content Translation module. |
AI accessed data may include information subject to GDPR, such as user contact information or any information subject to ticket services. User contact information may be supplied by external sources, such as Microsoft Entra ID.
Apart from the internal AI functions, Alvao can also be connected to general AI agents, such as Microsoft Copilot and others. For more information, see the ALVAO Microsoft Copilot Connector, ALVAO Asset Management MCP Server, and ALVAO Service Desk MCP Server modules.
AI-generated content is clearly marked in the application.
For example, titles of AI-generated knowledge articles start with the [AI] text.
External AI services
Alvao uses external online services to fulfill AI tasks.
For Foundry Models, customer data is not used to train or improve models. Alvao uses only the completion and embedding API calls that contain prompts. Fine-tuning and training files are not used. Data from prompts and completions can be temporarily stored by models in the same area for up to 30 days. This data is encrypted and accessible only to authorized Microsoft employees for the purposes of (1) debugging in the event of a failure and (2) investigating patterns of abuse to determine whether the service is being used in a manner that violates the applicable product terms. Microsoft hosts models in the Azure infrastructure, and all customer data submitted to models remains in the models service. Alvao uses the GPT-5.1 model.
For more information, see: