AI Field Agents let you embed AI-powered assistants into individual fields so that SmartSuite can analyze record context and generate structured results automatically, accelerating repeatable tasks and surfacing insights where work happens.
AI is built into SmartSuite’s platform, not bolted on — powering content generation, summaries, classification, and informed recommendations while maintaining governance, traceability, and enterprise security.
What Are AI Field Agents?
AI Field Agents are record-level AI assistants that:
Use custom prompts combined with data from other fields
Execute automatically based on field changes or manually on demand
Generate structured intelligence directly into SmartSuite records
Respect permissions and audit trails
Are governed with workspace-level controls
The AI output becomes the field value, stored with full history and logged as an AI event, enabling end-to-end traceability.
Supported Field Types
You can enable AI Field Agents on these fields:
SmartDoc
Text Area
Single Select
How AI Field Agents Work
You create a prompt that blends static text with dynamic field values.
SmartSuite automatically configures the Field Agent in the background using your prompt directions.
The agent runs:
Automatically when trigger fields change
Manually (user-initiated)
Results are saved into the field and logged as AI-generated.
This brings intelligence into workflows while preserving accuracy, governance, and auditability.
Workspace-Level Enablement (Admin Settings)
Only Workspace Admins can enable AI Field Agents.
Step by Step
Go to Workspace Administration > AI Control Center > AI Field Agents.
Turn on the Enable AI Field Agents toggle.
Optionally configure:
OpenAI API key for high-volume/paid usage
Internet Search support
View your Free AI requests remaining
Free AI Requests
SmartSuite includes a free request allowance so teams can start with AI without a paid key.
Works out of the box with limitations
When free requests are exhausted:
Agents remain enabled
Runs fail with an inline message
Adding your own API key unlocks continued usage
Free requests help teams trial and learn AI-powered workflows before they commit to API usage.
Enable Internet Search (Optional)
If allowed by Admins, AI Field Agents can supplement their context with internet search, expanding insights beyond record content.
Disabled by default
Enforced at workspace level
Controlled per field when available
This balances intelligence with governance and compliance requirements.
Configure an AI Field Agent (Field Level)
Once enabled workspace-wide, you configure AI Agent behavior per field.
Step by Step Configuration
Open Field Settings on a supported field.
Select the AI Field Agent tab.
Turn on Enable AI Field Agent.
Key Settings Explained
Prompt Builder
Mix static text and field references
Prompts must include at least one field variable
Prompt builds the context AI uses to generate useful outcomes
Example:
Based on {{Issue Description}} and {{Priority Level}}, summarize root causes and recommended next steps.
Avoid deleted or missing fields, as that can break the prompt.
Example prompt showing field selector:
Model Selector
Choose which LLM to use based on workspace configuration and capabilities.
Models may include:
OpenAI
GPT 5 Mini (Default)
GPT 5.1
GPT 5 Nano
GPT 5
GPT 4o Mini
GPT 4o Search
GPT 4.1 Mini
GPT 4.1
O3 Mini
O1 Pro
Note: SmartSuite continues to add platforms and models to the available list of LLMs. Check the list frequently to see the newest-available models.
Selection determines quality, cost, and response behavior.
Run Automatically
Enable this to let SmartSuite re-evaluate the AI Agent whenever referenced fields change, embedding intelligent updates directly into your workflow without human intervention.
Internet Search Toggle (Field Level)
Shows only when workspace Internet Search is enabled. This gives the AI external context beyond record data.
Note
The Internet Search toggle will only display when you have selected a model that supports Internet search and Internet Search Support is enabled in the admin console.
Manual vs Automatic Execution
Manual Runs
Users can invoke the agent directly:
From Grid view
From Record page
From a Record Details widget
Manual runs give control when reviewing or drafting content.
Automatic Runs
When enabled:
Runs are triggered behind the scenes
Results are applied and auditable
Runs occur only when relevant field values change
This lets workflows adapt intelligently as work progresses.
What Users See (States & Indicators)
Processing
Shows a spinner or progress indicator while AI runs are pending.
AI-Generated Value
When a run succeeds:
Value appears in the field
Activity shows Updated by AI
Field shows an indicator that output came from the agent
This supports audit and governance.
Manual Edit Attribution
If a user modifies the field:
Activity shows Updated by {User Name}
AI indicators are replaced
Error States
Possible errors include:
AI Agent Failed. The AI Agent is not enabled at the Workspace level, or there is a problem with the Field Agent's configuration.
No AI requests left. Your AI Platform token has exhausted available requests.
Invalid API key. The provided API key does not validate with the AI Platform provider.
Model error. The AI provider platform has returned an error.
Errors show inline with a red indicator and useful hover text.
Broken Prompt
If a referenced field is missing:
A message shows the prompt is broken
Field does not run
Auditing & Governance
Every AI Field Agent action is recorded in SmartSuite’s audit trail as AI-driven activity. This aligns with enterprise governance and audit requirements and supports visibility into who/what changed a field and why.
Duplicating Fields & Tables
Duplicating a field with an AI Agent preserves:
Prompt configuration
Underlying automation
Duplicating a table applies the AI configuration to new instances automatically.
API Key Management & Validation
When entering an API key:
SmartSuite validates it
Errors are shown up front
Invalid keys disable Agents until corrected
Admins are notified with a single digest email per failure cycle.
Disabling AI Field Agents
When an Admin turns off AI Field Agents:
All agents deactivate
Fields revert to standard behavior
Configurations are preserved
Confirmation messages ensure workspace admins act intentionally.
Best Practices
Include structured field references in prompts
Use automatic runs for dynamic workflows
Restrict Internet Search in compliance-sensitive environments
Monitor request usage and key health
Review AI output before applying to critical fields
Practical Use Cases
AI Field Agents help teams accelerate routine work while keeping intelligence embedded directly inside workflows.
GRC (Governance, Risk & Compliance)
Risk Description Summaries
Auto-generate concise risk narratives from data fields.
Control Test Conclusions
Summarize evidence and outcomes into concise result statements.
Regulatory Impact Insights
Summarize guidelines and impacts (with Internet Search enabled).
ITSM (IT Service Management)
Incident Summary Generation
Turn verbose incident text into structured summaries.
Root Cause Drafts
Generate first-pass root cause analysis from fields like Impact, Timeline, and Resolution Notes.
Change Risk Assessments
Provide AI-backed recommendations based on related data fields.
Summary
AI Field Agents embed intent-driven intelligence into your core workflows on a single, unified SmartSuite platform — empowering teams in GRC, ITSM, and beyond to work faster, make smarter decisions, and stay aligned with organizational policies and audit requirements.