Exploring Flow recommendations
Summarize
Summary of Exploring Flow recommendations
ServiceNow’s Flow recommendations feature in Workflow Studio leverages an AI-driven model to suggest the next component in your flow based on the current flow position and previously used components. These AI-generated recommendations help streamline flow creation by providing relevant next steps, improving efficiency and accuracy as you design workflows.
Show less
Key Features
- AI-Generated Suggestions: The system analyzes the names of prior flow components to generate one to five recommendations for the next step, including actions, flow logic, and subflows available from ServiceNow.
- Scope of Recommendations: Recommendations exclude custom user-generated components and actions from ServiceNow Store spokes, focusing solely on standard ServiceNow components.
- Generative AI Model: The underlying large language model was trained on internal ServiceNow flows to recognize relevant flow creation patterns and suggest appropriate next components.
- User Control: Users can toggle the display of flow recommendations on or off on a per-flow basis via the “Show recommendations” flow preference setting.
- Consistent Experience Across Views: Whether editing in diagramming or text view, recommendations appear at the end node of the flow, maintaining a consistent user experience.
Important Considerations
- Limitations of AI Recommendations: As the AI predictions are based on learned patterns, they may not always be accurate, complete, or fully appropriate for all use cases.
- Responsibility and Oversight: Customers should thoroughly test and evaluate AI-generated recommendations for accuracy and relevance, applying human oversight and not relying solely on AI outputs for critical decisions.
- Compliance: Usage of this feature must comply with ServiceNow’s AI Acceptable Use Policy, which governs responsible use and may be updated over time.
Select the next component in your flow from a list of AI-generated recommendations. The system generates recommendations based on the current position in the flow and the flow component names listed before.
The model uses the name of the flow components that come before to generate one to five recommendations for the next step of the flow. If there are no recommendations listed, then there are no flow components that meet the required relevance threshold.
The system can only recommend actions, flow logic, and subflows that are available from ServiceNow. Recommendations can’t include user-generated flow components such as custom actions, nor can recommendations include actions from ServiceNow Store spokes.
Generative AI model training
This Generative AI large language model was pre-trained with internal ServiceNow flows to learn flow creation patterns. The goal was to understand what flow components are most relevant for a certain position in a flow given the content before.
Flow preference
By default, Workflow Studio shows flow recommendations as you build a flow. You can hide these recommendations on a flow by flow basis by turning off the Show recommendations flow preference. See User preferences for flows for more information.
Recommendations in diagramming view
Workflow Studio shows recommendations from the flow diagramming view while you're editing a flow. In diagramming view, the last node of the flow always shows a list of recommendations. This experience matches the behavior of the text view, which also always shows a list of recommendations at the end of the flow.
AI limitations
This application uses artificial intelligence (AI) and machine learning, which are rapidly evolving fields of study that generate predictions based on patterns in data. As a result, this application may not always produce accurate, complete, or appropriate information. Furthermore, there is no guarantee that this application has been fully trained or tested for your use case. To mitigate these issues, it is your responsibility to test and evaluate your use of this application for accuracy, harm, and appropriateness for your use case, employ human oversight of output, and refrain from relying solely on AI-generated outputs for decision-making purposes. This is especially important if you choose to deploy this application in areas with consequential impacts such as healthcare, finance, legal, employment, security, or infrastructure. You agree to abide by ServiceNow’s AI Acceptable Use Policy, which may be updated by ServiceNow.