This article will provide more information on all the Agentic AI capabilities that have been released by ServiceNow.
May 2025 Release (current release):
The May 2025 Store Release has two agentic capabilities: Resolve HR Cases, AI Agents for Growth Conversations.
Resolve HR Cases
For the "Resolve HR Cases" use case, we have shipped 4 agents Out of the box to work – Record management agent, HR criticality detection AI agent, HR search retrieval AI agent and HR notification Agent. This agentic workflow automates the resolution of employee queries by leveraging AI agents to create case, detect "criticality", and shares relevant KBs/ catalogs with the employee.
When an employee submits their question through email or on the portal, the AI Agents will gather the case details and analyze if the case is critical. If the case is critical, then the case is routed to a human agent to service. If the case is analyzed to be noncritical, then the HR search retrieval AI Agent surfaces up the relevant KBs and catalog items that could resolve the employee’s question. At the end, the notification agent then updates the case and sends a notification either via email or through Now Assist in Virtual Agent.
AI Agents: Growth Conversation Planning
In the Agentic AI space, two agents are being released specifically to assist managers in planning growth conversations with their employees.
- The first agent is the conversation management agent, which collects employee details, checks for upcoming conversations(if any), and schedules new conversations by finding common time slots, including those in the Outlook calendar.
- The second agent is the employee data summarization agent, which compiles and summarizes employee’s activities, such as skills, learning, feedback, growth plans, journeys, mentoring activities, and more. It also suggests talking points for the conversation and adds them to the conversation agenda. This agent is triggered subsequently after the first agent, subject to the initial request message.
Manager can trigger these agent by initiating conversation with Now Assist Virtual Agent chat on employee portal. These agents aim to streamline the process of managing and conducting growth conversations, making it easier for managers to engage with their employees effectively.
March 2025:
In March 2025, we released capabilities such as AI Agent Studio, Help Resolve Tuition Requests, as well as AI Agents for High Volume HR Requests.
AI Agent Studio
To deploy and configure AI Agents, you can use the AI Agent Studio to help you get started. The AI Agent Studio guides you through the deployment process so you can automate and realize value quickly.
Before diving deeper into Agentic AI, let’s define some new terms, like use case and AI Agents. A use case is the business objective and an end goal we are trying to reach. Each use case will have various AI Agents working together. These AI Agents will have roles and instructions on what they need to do. There are various platform tools like workflows, integrations, and skills aligned to the AI Agents. These tools allow the AI Agents to autonomously take action on the tasks that they are assigned to.
In the AI Agent Studio, there are various capabilities to make it easy for you to implement AI Agents:
- Prompt Rewriting – Auto-generation of instructions for AI agents and use cases to mitigate prompt paralysis + streamline setup/config
- Recommended Tools/Agents – System-recommended Tools and AI agents to mitigate the need to navigate to other apps
- OOB BU Agents and Use Cases – Promoted OOB Agents and Use Cases
- Overview Video – Conceptual breakdown of AI Agents
- Journey Checklist – checklist for setting up and configuring AI agents/use cases
On the “Create and Manage” tab, this is where you can easily activate and configure your AI Agents. Here you can modify the AI Agents, as well as provide tools and instructions on the AI Agent’s role and responsibilities.
After configuring the AI Agents, you can test them out right from the Studio view and quickly edit them in one place.
Along with these capabilities, there is also a dashboard available to you to measure value and success.
- Analytics – Dashboard built in Performance Analytics to measure KPIs for impact on business (tasks resolved using AI agents, conversations using AI agents, etc) as well as usage metrics (execution plans, AI agents tasks, etc)
With the AI Agent Studio, we are providing one place to get all configuration work done, so you have an easy and fast deployment experience.
Help Resolve Tuition Requests
In our March release of AI Agents, we are releasing an out of the box use case for requests that require policy validation, which are usually manual and require high effort for resolution.
In this case, this is built specifically for tuition reimbursement scenarios where you need to reference a policy to approve or disapprove a request that comes in. For this use case, there are seven out of the box agents that are mapped to it. These AI agents will accelerate the way human agents can resolve cases.
Team of AI Agents for this use case:
-
The seven agents include:
- HR total rewards - case details manager: This agent can fetch details of a HR Total Rewards case.
- Tuition reimbursement requestor information retriever: This tool can collect details about the tuition reimbursement requestor
- Tuition reimbursement policy content retriever: This agent can fetch the content of tuition reimbursement policy.
- Tuition reimbursement checklist creator: This agent specializes in generating a detailed checklist of criteria that must be satisfied to obtain approval for tuition reimbursement request. This agent can execute only if Tuition Reimbursement policy content is provided.
- Tuition reimbursement eligibility evaluator: This agent can assess tuition reimbursement requester eligibility for tuition reimbursement pre-approval.
- Approval task creator: The Approval Task Creator agent is responsible for creating new approval tasks. It first checks for any existing approval tasks, notifies the manager about them, and proceeds to create a new task if none exist.
- HR case closure evaluator: This agent can update HR case state and work notes for the case closure.
Key configurations:
For this use case, you can start by going to your AI Agent Studio. Here, you will provide instructions in your natural language on what the end goal for this use case is.
From there, you can define what AI Agents are needed for this use case.
These AI Agents can be further defined with instructions and the tools that are needed to execute on their assigned work.
After you’ve set up the AI Agents and the use case, you can test out the AI Agents using the Testing feature. You can see the individual AI Agents getting activated and see them in action. You can make changes right away and ensure you are creating the desired experience.
AI Agents for High Volume HR Requests
Alongside with the policy issue resolver use case, we released standalone AI Agents that can be used to address high volume HR requests.
Here are the 8 AI Agents that are included with this release:
- Employee time off request assistance AI agent
- Employee holiday calendar retrieval AI agent
- Employee total rewards information AI agent
- Employee profile information AI agent
- Expense details completion AI agent
- Employee personal information updates AI agent
- Employee performance feedback AI agent
- Employee benefits information AI agent
These 8 AI Agents can be used to configure a larger use case and help resolve cases faster for HR teams. Note that these standalone agents call on existing integrations that can be used in resolution for the listed HR queries.
Key Configurations:
To get started, you can access these individual AI Agents in your AI Agent Studio. From there, you can provide instructions on what the AI Agent’s role and responsibilities will be.
From there, you can add tools, like integrations to other systems, to pull in the PTO information to ServiceNow. Once you have completed this step, you can click on Testing and see your AI Agents in action.
We'll update this article once a new release is live, so check back here for updates!
2 Comments
