Applaud AI helps employees find information, complete HR tasks, and get HR support. Employees can use natural language to ask questions, retrieve live information, take approved actions in connected systems, start workflows, and raise HR cases.
HR teams can also use AI to support case intake, triage, summaries, drafting, and other case-management tasks.
What employees can do depends on the knowledge, tools, agents, workflows, permissions, and channels you configure.
Your organization's data is not used to train foundation models. Applaud-hosted AI inference runs through Amazon Bedrock in your tenant's data-residency region.
What the AI assistant can do
The AI assistant is the main way employees interact with Applaud AI. Depending on your configuration, it can:
- Answer questions using indexed organizational knowledge, including policies, how-to articles, FAQs, and connected SharePoint content.
- Retrieve live information from connected HR systems.
- Take approved actions using Agent Tools, such as submitting requests or generating documents.
- Delegate work to specialist agents that handle specific domains, countries, or processes.
- Start and continue workflows for designed, multistep processes.
- Support HR cases with AI-assisted intake, triage, summaries, drafting, and case planning.
In the employee portal and mobile app, the assistant can also present information using Applaud visuals, such as cards, lists, charts, maps, and org charts. In Microsoft Teams, Slack, and Outlook, these visuals appear as images.
Where employees use the AI assistant
The same AI assistant can be made available through several channels, depending on your plan and configuration.
| Channel | How employees use it |
| Employee portal | Chat with the assistant in the Applaud experience |
| Mobile app | Chat with the assistant in the Applaud app |
| Microsoft Teams | Message the Applaud assistant |
| Slack | Use a thread, direct message, or @mention |
| Outlook | Email a connected shared inbox and continue the conversation in the same thread |
Slack, Teams, and Outlook are available with Standard and higher plans under current packaging. For channel availability and differences, see AI support across channels.
Some pages can also provide an instant answer without creating a conversation in history. Selecting Continue in chat turns the interaction into a normal conversation.
Knowledge: answers from your content
The AI assistant uses your organization's content to answer questions about policies, processes, and other HR topics.
You can provide knowledge through:
- Knowledge Management articles, including policies, how-to articles, and FAQs.
- Knowledge documents, such as PDF, Word, Excel, and other supported files.
- SharePoint, including connected document libraries, site pages, and lists.
The assistant searches indexed knowledge. Files that employees attach to a conversation can be used during that conversation, but are not added to the shared knowledge base.
Tools and workflows: AI that can act
Knowledge helps the assistant answer questions. Agent Tools allow it to retrieve live information and take supported actions in connected systems.
Experience Designers can connect systems such as Workday, SAP SuccessFactors, custom APIs, and remote MCP servers, then make specific capabilities available to the assistant as tools.
Each tool has a low, medium, or high risk level. Higher-risk actions can require additional authentication. For actions that change records, the assistant can ask the employee to review and confirm the request before submitting it.
Agent Tools can be used across:
- Everyday conversations
- Case Management
- Workflow Agent and Run Tool steps
- Supported external AI clients.
Agentic workflows combine AI with defined steps, integrations, waits, and conditions. This lets you design processes such as leave requests and onboarding journeys with a defined path and outcome.
Connect external AI clients
Applaud can connect with external AI clients through MCP (Model Context Protocol).
Inbound MCP enables supported clients, such as Microsoft Copilot, ChatGPT, Claude, and Gemini, to connect to your Applaud tenant. When configured, employees can access Applaud capabilities through the external client using their Applaud identity and the knowledge and tools available to them.
Outbound MCP allows you to make tools from another MCP server available in your Applaud Agent Tools catalog.
AI in HR Case Management
When Case Management is enabled, AI supports HR agents rather than replacing them.
AI can help with:
- Intake: Clarify the request, identify the appropriate category, and capture information needed for intake.
- Triage: Suggest a team, assignee, category, flag, or priority.
- Summaries: Create a plain-language summary of the case.
- Drafting: Help agents draft responses privately in the case composer.
- Case plans: Create an AI-generated task list for cases that require multiple steps.
- Experience metrics: Provide AI-generated evaluations that contribute to case quality measures.
AI remains in a supporting role. HR agents can change triage information, edit or reassign case-plan tasks, override AI-generated experience metrics, and decide whether to send a drafted response.
If the AI cannot determine the appropriate triage outcome with sufficient confidence, the case can be flagged for manual triage. Public replies are sent by HR agents rather than automatically posted by the AI assistant.
Employees see the public conversation, not internal AI activity or case plans.
Languages and translation
Applaud supports both product localization and case message translation.
Product languages
The Applaud interface is available in multiple locales, including right-to-left layouts. Conversation starters and knowledge content can also be localized.
The AI assistant can respond in the employee's language.
Case message translation
When a case involves people who use different languages, Case Management can translate messages in the conversation.
When another language is detected, a translation option is available in the case. Once enabled, later incoming messages in that case can be translated automatically.
Case translation is intended for conversations between people who do not share a language. It is not intended to translate an entire knowledge library.
Trust, privacy, and compliance
AI capabilities include controls designed to help organizations manage AI responsibly.
Your data and AI models
Your data is not used to train foundation models.
Customer prompts, knowledge, conversations, and case content are not used to train foundation models.
Applaud-hosted AI inference runs through Amazon Bedrock.
AI inference for the Applaud assistant runs through Amazon Bedrock in the region where your tenant is deployed.
AI processing follows your tenant's data residency.
AI traffic is processed in your tenant's residency region. For example, an EU tenant uses EU Bedrock endpoints.
The foundation models available to Applaud are provided through Amazon Bedrock. Applaud can select models for different AI capabilities, such as conversations, summaries, and classification.
For information about subprocessors and contractual data-processing terms, refer to your organization's applicable legal and DPA documentation.
Oversight and risk controls
- Human oversight: AI-supported actions and recommendations can generally be reviewed or changed by a person before they are completed or shared.
- Risk levels: Every Agent Tool has a low, medium, or high risk level. Higher-risk actions can require additional authentication.
- Confirmation: Supported actions that change records can require the employee to review and confirm the action before it is submitted.
- Guardrails: Messages and attachments are checked for risks such as sensitive content and prompt injection.
- Citations: Knowledge-based answers can provide citations to the content used to generate the response.
- Audit information: AI activity and case evaluations are recorded so authorized users can review relevant information.
- Usage limits: AI usage is measured against your plan, including applicable message, token, attachment, case translation, and case-planning limits.
Supporting your compliance obligations
Applaud provides controls to help organizations meet obligations under GDPR, the EU AI Act, and similar regulations. These include purpose-limited processing, regional data residency, no training on customer data, human oversight of consequential decisions, risk classification for AI actions, and an evidence trail for AI-supported decisions and evaluations.
Your organization is responsible for configuring knowledge, tools, cases, access, and data retention to meet its own policies and regulatory requirements, including any applicable DPIAs.