Liz AI assistant in Trento Web

Liz is a chat-based AI assistant built directly into Trento Web. It lets you ask questions about your SAP landscape in natural language. By analyzing the live state of your systems and best-practice validations, Liz helps you actively monitor health and identify issues, all without requiring any additional deployment.

This chapter covers the following topics:

About the Liz AI assistant

Liz is part of Trento Web itself and requires no additional deployment. You only need to supply an API key for a supported large language model (LLM) provider.

Liz is guided by built-in system instructions tailored for Trento, ensuring it understands your SAP landscape right out of the box.

Data privacy

Liz sends your prompts and the system data needed to answer them to the third-party LLM provider you configure. Review your organization’s data handling policy before enabling this feature.
Which providers and models are supported?

Liz AI assistant currently supports the following LLM providers:

  • Anthropic Claude

  • Google Gemini

  • OpenAI GPT

Each provider offers several models. To ensure compatibility, Trento only displays models capable of powering the assistant.

How does Liz access data?

Liz answers queries using live SAP landscape state from Trento Web, and check results from Trento Wanda.

Access is bound to your Trento user account. When you open a chat session, Liz can only access data you are authorized to see. Your existing Trento roles and permissions remain unchanged.

Configuring Liz

You can configure Liz from your user profile in Trento Web.

Requirements:

  • A user account in Trento Web. AI configuration is per user and each user configures their own provider and model.

  • An API key for one of the supported providers.

To configure Liz:

  1. In the Trento Web console, click your user icon and select Profile to open your profile page.

  2. Go to the AI Configuration section and click Edit Settings.

  3. Select a provider and model from the supported list.

  4. Enter your API key for the chosen provider.

  5. Click Save to save your configuration.

Liz becomes available as soon as the configuration is saved.

You can change your model or provider at any time from the Profile page without the need to reload the web console. Changes apply starting with your next prompt; any prompt currently processing finishes using the previous configuration.

To stop using Liz, clear your AI configuration from the Profile page. If the chat window is open, Liz goes offline and the conversation becomes read-only. Once the window is closed, the chat launcher is disabled.

Using Liz

Open the chat launcher in Trento Web and type your question. Your chat session remains active as long as you are logged into the Trento Web console.

Example queries:

  • What is the tuning status of the registered hosts?

  • Which hosts are affected by patch XYZ?

  • Provide a summary of checks not passing for cluster XYZ and step-by-step remediation instructions.

If you did not configure a provider and model, the launcher is disabled. See Configuring Liz for more information.

While a conversation is open you can perform the following actions:

Action Description

Stop

Stop Liz processing the current question. The assistant stays available for your next question.

Retry

Re-send your last prompt.

Copy

Copy a reply to the clipboard. Liz formats its responses in Markdown, so pasting into compatible tools (like wikis or issue trackers) preserves the text formatting.

Selecting Learn more in the assistant opens the Trento documentation in a new browser tab.

Clicking New chat in Liz immediately ends the active conversation and clears its history. Previous chats cannot be recovered once a new session starts.