Sage People AI FAQs

Important This content is part of the Sage HR Assistant Early Adopter release. If we haven't contacted you to be part of this program, refer to future release information.

Read our frequently asked questions and how to solve them.

What is the data model for HR Assistant?

HR Assistant combines structured data (Team Member, Employment etc.) from your Salesforce enterprise org with unstructured data from your Policies.

What is the data model for Performance Coach (past review summaries and development objective draft)?

Performance coach uses data you've stored in your Sage People org.

Will AI draw from external systems, for example when creating objectives?

No. Sage People AI will only use information within the platform and won't access information from external sources like the web.

is there a feedback loop in the system?

This is available natively for HR Managers using agents you're presented in the Salesforce AI conversation panel. Feedback for performance review summaries and draft objectives is on our roadmap for EA.

What data does Agentforce access (for example, configuration data, employee data)?

Agentforce only uses your organization data, including Sage People records, HR data, uploaded documents, and configured knowledge libraries. It retrieves this data in real time using secure access controls.

Agentforce Data Libraries governs data access. You configure these as part of the agent setup and scope them to include only the data you require for each function.

For example, the HR Assistant has access to team member data to answer employee-specific queries, and policy documents you supply. This is configurable to control the data it can access. It will evolve throughout the development of Sage People AI and we'll keep you up to date.

Will customer data remain secure and isolated?

Yes. The Salesforce platform hold your data securely. We protect communication with external LLMs with the Einstein trust layer protects and by contractual arrangement with LLM providers. This ensures the LLM provider doesn't retain data or use it to train its models.

Is customer data used in any global model training?

No. Salesforce have contractual arrangement with LLM providers to ensure they don't retain data or use it to train its models.

How will it work with GDPR and right to be forgotten?

The agent itself holds no data, so the data deletion processes will continue as per your set up now. LLMs don't use any employee data to train its models.

Where does the system store the source data?

The system stores all data within your own Sage People org or Salesforce Data Cloud instance. You retain control over how long you keep it and when you delete it. Discuss your individual requirements during implementation.

See the Salesforce articles on how to find the tenant and where it's located.

What is Sage People's policy regarding AI development and use?

Read out article on responsible AI built for you.

Which LLMs are we using?

  • Performance Coach Summaries - Anthropic Claude Sonnet 4.5 on Amazon

  • Performance Coash Objectives - Anthropic Claude Sonnet 4.5 on Amazon

  • HR Assistant - Anthropic Claude Sonnet 4.5 on Amazon

  • Policy Insights - GPT5

Is there guidance on AI security from Salesforce?

Yes, see the Salesforce article on AI agent security.

How can I ensure transparency and explainability of Sage People's AI systems' decisions and outputs to end-users?

For HR Assistant, the audit trail will display answers it gives and the data sources it used to get these answers.

For Performance Coach, this is currently under development and we'll deploy it before general availability.

What is the release cycle?

We'll do several releases throughout the early adopter period. We'll then deploy changes in line with the regular Sage People release cadence.

How can customers give feedback?

During the early adopter period, our implementation and design teams are working closely with customers to gather feedback. Once we move to general availability, we'll gather feedback through Executive Business Reviews and the customer ideas hub.

Who can I ask for help?

When you join the early adopter program, we set you up with an implementation team. They can answer your questions and help set up your platform.

You can ask your customer success manager any questions you have on billing or request additional credits.

Once your implementation is complete, you'll raise any technical questions through the support team via community.

Can I disable the AI tool if I need to?

Yes, the AI tools are fully in your control. You can adjust certain features and disable to platform in full at any time.

What are credits?

We include credits as part of your Sage People with AI agreement. They represent units of work that the AI agent performs.

Each action the agent carries out consumes credits. Examples include:

  • Retrieving data (for example HR records, cases, or documents)

  • Updating a record

  • Running a workflow

  • Generating a response or summary

What does free AI capacity mean?

As part of your Sage People with AI license, you receive numerous free credits per year. If your usage will exceed this amount, you can buy additional credit bundles to support higher levels of AI activity.