Overview:
Generate My Work Plan helps you stay organized by creating a personalized schedule based on your assigned tasks. It considers urgency, deadlines, impact, sentiment, and SLAs, then estimates how long each task will take and when you can finish it. The result is a clear, easy-to-follow plan that helps you focus on what matters most.
What Problem Is It Solving?
Organizations frequently encounter challenges in efficiently planning work and ensuring teams are adequately prepared before execution. In the absence of AI-driven agents, these processes remain largely manual, slow, and susceptible to errors. Individual agents are often required to personally review their assigned tasks, assess priorities, estimate the effort required, and develop daily work plans, resulting in significant planning overhead. This manual approach is not only time-consuming but also inconsistent across agents, which can lead to suboptimal task sequencing and an increased likelihood of missed service level agreements (SLAs).
Examples:
-
A service desk agent asks, “What should I work on today?” and receives a prioritized list of incidents with estimated resolution times.
-
A field technician says, “Plan my work,” and gets a schedule that highlights which tasks are most urgent and which ones are blocked.
-
A manager uses the workflow to understand which backlog items are closest to breaching SLAs and should be tackled first.
Setup Requirements:
-
License required – Professional or Enterprise+
-
Plugin Version– Minimum patch required Yokohama Patch 6 or Zurich Patch 1
-
Plugins – Must have Now Assist for Platform (included with apps like Now Assist for ITSM, CSM).
Below screen shot is an example of Now Assist for ITSM which has also activated the dependency plugin Now Assist for Platform.
Required User Roles:
-
'sn_uxc_gen_ai.platform_ai_work_planner' role is required for this agentic workflow.
-
'sn.now_assist_panel_user' role is required to run the workflow from the Now Assist Panel. Additionally, ensure that the system property 'sn_ais_assist.dpr_ingestion_completed' is set to true.
-
You can also run this workflow on the Testing page of AI Agent Studio with the same utterance/prompt in the Task field if you have the 'sn.aia_admin role' .
Activation:
Triggers can be enabled for the Now Assist Panel to allow users to initiate workflows with natural language commands.
Go to All -> AI Agent Studio -> Create and Manage -> search for ‘Generate my work plan’ -> click on it and go to ‘Select UI display’ to activate the display toggle for Now Assist panel -> and then click on ‘Save and test’ as shown in the below screenshot.
-
Sample Utterances or Prompts: ‘What should I work on today?’ or ‘Plan my’ or any similar phrases in the Now Assist panel to trigger the workflow.
-
When an utterance is provided, the Now Assist Panel recognizes the respective agentic workflow and triggers the respective agentic workflow orchestrator to execute the next steps.
How Does It Work?
Instead of sifting through tasks manually, the Generate My Work Plan agentic workflow—powered by AI agents—reviews the work assigned to you, checks dependencies, flags potential blockers, and helps you prioritize effectively. This ensures that your personal work plan is clear, achievable, and aligned with deadlines.
Below is an example screenshot of the agentic workflow results, providing the final work plan for the user to focus on high-priority tasks before moving on to the next level of tasks. Each task includes task information, an estimated start date, an estimated completion date, and the estimated number of hours required to complete the task based on similar historical records. It also provides reasoning for why a task is considered high or low priority and includes comments for any missing information or cases where there are no similar records to provide estimations.
How do the AI agents work together to provide the results?
When the "Generate My Work Plan" agentic workflow is triggered—either through natural language commands or the Now Assist panel—the workflow orchestrates three AI agents to automatically scan all assigned tasks across any workspace, list, or queue and generate a prioritized work plan. This process helps fulfillers focus on the most impactful work, reduce SLA breaches, and maximize success metrics.
- Prioritize Work AI Agent:
-
Retrieves all tasks assigned to the logged-in user.
-
Calculates a combined score for each task using:
-
Task priority
-
SLA urgency
-
Sentiment
-
-
Sorts tasks by score into a list called prioritized_tasks.
The screenshot below displays the list of tasks in order, along with their scores, to prioritize the list.
- Estimate Effort AI agent:
Loops through each task in prioritized_tasks.
For each task:
-
Searches for historical or related records.
-
If data is sufficient:
-
Estimates effort in hours.
-
Provides a one-line rationale.
-
If data is insufficient:
-
Returns a message about lack of data or no similar records found and a Note comment.
-
Examples:
-
Ensures every task has either an estimate or a clear fallback message. If any task is missing an estimate, it reattempts estimation before moving forward.
- Plan Work AI Agent
-
Uses effort estimates to calculate realistic completion dates.
-
For tasks without estimates:
-
Includes them in the plan with a note about missing data.
-
Suggests how users might obtain the missing info.
-
Builds a user-centric, actionable work plan by producing a structured plan with estimated hours per task. Outputs a clear sequence of what to start first and why.
-
Output Formatting: Presents the final work plan as a bulleted or numbered list.
-
For each task, it includes:
-
Task name or description
-
Priority
-
Estimated effort (or fallback message)
-
Estimated completion date (or note if unavailable)
-
Estimated total hours to complete the task
-
Reasoning for the priority.
-
-
These AI agents work together to ensure the resulting plan is clear, user-friendly, actionable, deadline-aware, and optimized for success metrics.
Demo
For the Demo – Watch the Platform Academy session - What's new in Zurich from 2:40 to 19:20.
Platform OOTB Agentic Workflows:
2:44 - Zurich released list of Agentic Workflows Overview
3:39 - Demo: Image-to-Incident Agentic Workflow
11:01 - Demo: Generate My Work Plan Agentic Workflow
Outcomes & Benefits
-
Efficient Work Planning and Prioritization: AI-driven tools empower employees to prioritize and organize tasks, creating dynamic work plans based on urgency, priority, and historical effort for optimized scheduling and fewer missed deadlines.
-
Productivity and Focus: Automated task triage and streamlined interfaces reduce administrative overhead, freeing up time for problem-solving and enhancing overall agent productivity.
-
Improved Outcomes and SLA Performance: Context-aware insights and prioritized planning minimize overdue tickets and escalations, resulting in better customer satisfaction and greater trust.
-
User Experience and Adoption: Intuitive, natural language interfaces make automation more accessible, driving user adoption and providing a smoother experience.
-
Customization and Expansion: Out-of-the-box workflows can be extended or customized, allowing organizations to build agentic solutions tailored to their unique needs and risk factors.
Limitations:
ServiceNow supports four out-of-the-box LLMs—Now LLM, Anthropic, Gemini, and GPT. However, for the September 2025 release, this agentic workflow may not be supported by Now LLM and might be considered for inclusion later. Please follow up with your account executives to find support for future releases.
Personas:
Fulfiller/ Agent
Additional Resources:
Explore platform product resources here, and review the list of helpful resources under Platform Agentic Workflows and AI Agents.
