Version
Follow us on YouTubeFollow us on YouTube (opens in a new tab)Contact us via SlackContact us via Slack (opens in a new tab)Go to Sema4.ai home pageGo to Sema4.ai home page
  • GET STARTED
  • Enterprise Edition
    • Build your first AI agent
    • Build agents with Custom Actions
    • Build agents with Named Queries
    • Publish agents to Control Room
    • AI Operator: Deploy agents
    • Interact with agents
      • Meet Work Room
      • Chat with Conversational Agents
      • Work with Worker Agents
      • Communicate Effectively with AI Agents
  • Team Edition
    • Install Snowflake native app
      • Installation
      • Try a demo agent
      • Admin guide
      • Multiple instances
    • Install Studio and SDK
      • Partner OAuth
      • Custom OAuth
      • Key pair auth
    • Agent templates
      • Cortex Analyst Example
      • Cortex Search Example
      • Sales Contact Finder
      • Similar Company Finder
    • Build your own agent
      • Data setup
      • Quickstart
      • Step-by-step walkthrough
    • Public API
    • Deploy agents
    • Work with agents
  • Release Notes
    • SDK
    • Studio
    • Control Room
    • Agent Compute
    • Python Libraries
    • Actions Gallery
    • Team Edition
  • DOCUMENTATION
    • Installing Studio
    • Working with SAI
    • Configurations
    • Shared Configurations
    • Agent Project Folders
    • Auto-suspend Agents
    • Agent Observability
    • Agent Evaluations
    • Conversation Guide
    • Conversation Starter
  • Build agents
    • Conversational Agents
      • Set up Studio
      • Agent templates
        • Set up and deploy the agent
        • Interact with the agent
      • Create your own
      • Publish your agent
      • Introduction to Worker Agents
      • Create a Worker Agent
      • Publish and Deploy Worker Agents
      • Add Work Items via API
    • Reasoning Models
      • Setup Model Providers
      • Model Compatibility & Benchmarks
      • Troubleshooting
    • Actions Gallery
      • Use with agents
      • Authentication
        • Set up OAuth2 clients
        • Github
        • Google
        • Hubspot
        • Microsoft
        • Slack
        • Zendesk
    • Model Context Protocol (MCP)
      • Docker MCP Gateway
      • Remote MCP Servers
      • Build Custom MCPs
      • Agent as MCP Server
    • Develop AI actions
      • Install SDK
      • Action templates
      • Action structure
      • Create custom action
      • Actions with data
      • Publish to Studio
    • Evaluating your Agent
    • FAQ
  • Manage agents
    • Meet Control Room
      • Organization
      • Agent Compute
      • Agent lifecycle
      • Get started
    • Publish to Control Room
      • Agent Gallery
      • Publish from Studio
      • Import agent packages
      • Manage agent versions
    • Deploy agents
      • Prepare environment
      • Agent deployment
    • Operate agents
      • Monitor agents
      • Update agents
      • Observability
      • Retire agents
      • Advanced Configuration
    • Audit Logs
    • FAQ
  • Work with agents
    • Meet Work Room
    • Chat with Agents
    • Work with Agents
    • Access data
    • Communicate with agents
    • FAQ
  • Channels
    • Microsoft Teams
      • Admin Guide
      • User Guide
    • Slack
      • Admin Guide
      • User Guide
  • Data Intelligence
    • Semantic Data Models
      • Databases
      • Structured Files
      • Import Data Model
      • Verified Queries
      • Editing Data Models
      • Deploying Agents
    • Dynamic Data Access
      • Local development
      • Actions with data access
      • Build agents with data
      • Deploy agents with data
      • About SQL dialect
      • Known limitations
    • DataFrames
      • Native Queries
    • Knowledge Base
      • Getting started with Knowledge Bases
      • Build a Knowledge Base
      • Define Queries for the Knowledge Base
      • Use Knowledge Base in an Agent
      • Deploy Agent with Knowledge Base
      • Maintaining Knowledge Base
    • Document Intelligence
      • AI-Guided Configuration
      • Agentic Processing
      • Agent Integration
      • Automatic Adaptation
  • API
  • Agent API
  • Work Item API
  • SUPPORT
    • Network requirements
    • Configure proxy setting
  • Terminology
  • Contact us
    • GET STARTED
    • Explore the demo agents
    • Quick-build with AI
    • Build an agent from scratch
    • Release Notes
    • DOCUMENTATION
    • Build agents
      • Agent types
      • LLM models and providers
      • Runbooks
      • Versioning & lifecycle
      • Tools with MCP
      • Integrations
        • Google
      • Data connections
      • Semantic data models
      • Data frames
      • Document Intelligence
      • Files
      • Advanced settings
      • Conversation guide
      • Evals
    • Work with agents
      • Versioning
      • Chat vs. work items
      • Tools that need authentication
      • Sharing threads
    • Administration
      • LLMs
      • User management
      • MCP servers
      • Branding
      • Observability
      • Advanced settings
    • Insights & audit
      • Analytics
      • Audit log
    • REST API
      • Agents
      • Conversations
      • Work items
      • Schedules
      • Data connections
      • Semantic data models
      • MCP servers
      • LLMs
      • Branding
      • Advanced settings
      • API keys
      • Users
      • OIDC role mapping
      • Audit log
      • Observability
      • Document Intelligence
      • Agent MCP
    • DEPLOY
    • Self-host Sema4.ai
      • Deployment targets
      • AWS EKS
      • AWS VM
      • Azure AKS
      • Azure VM
      • Reference
      • Identity providers (OIDC)
        • Microsoft Entra ID
        • Auth0
        • Okta
      • Network endpoints
      • Configure LLMs
      • MCP servers
      • Set up OAuth for MCPs
      • Set up MCP orchestration
        • Google Cloud Run
        • AWS ECS Fargate
        • Azure Container Apps
        • Snowflake SPCS
      • Configure workspace branding
      • Set up Observability
      • Update the Sema4.ai application
      • Add networking rules
      • Add LLM configurations
    • Snowflake Native App
      • Installation
      • Snowflake Cortex LLMs
      • User access and roles
      • Data connections
      • API access
      • Maintenance
      • Logs and Telemetry
    • SUPPORT
    • Reporting issues
    • GET STARTED
    • Get started
      • How the three personas work together
      • Quickstart for IT
      • Quickstart for builders
      • Quickstart for analysts
    • Release notes
    • DOCUMENTATION
    • Administration
      • Workspace, teams, and users
      • Single sign-on and identity
      • Connections
        • How connections get set up
        • Integrations
          • Capabilities
          • Catalog
          • Credentials and ownership
          • Sync status and troubleshooting
          • PROVIDERS
          • SharePoint
          • Microsoft Teams
          • OneDrive
          • Outlook Mail
          • Outlook Calendar
          • Google Drive
          • Gmail
          • Google Calendar
          • GitHub
          • Slack
          • Linear
          • Ramp
          • Stripe
          • Exa
        • MCP servers
        • Data sources
          • Snowflake
          • Connect non-Snowflake providers
          • Postgres
          • Snowflake Profiles
      • Configuration
      • Notification defaults
      • Audit and evidence
        • The dual-lens model
        • Deterministic execution
        • Safe execution and isolation
        • COSO alignment and controls
        • Export, retention, and evidence packages
    • Build
      • Projects
        • Create a Project
        • Author Instructions
        • References and attachments
        • Test and iterate
      • Shared context
        • Shared memories
        • Verified logic
        • Shared versus project scope
      • Project context
        • Files and the Project VFS
        • Project Memories
        • Skills
        • MCP mounts
        • Integrations
        • Attached Data Collections
        • Email templates
        • Schemas
      • Apps
        • Create an App
        • Author the Runbook
        • Inputs, outputs, and work item contracts
        • Triggers
        • Bound email
        • Review and publish
        • Versions and rollback
      • Data Collections
        • Create a Data Collection
        • Add data sources
        • Scope the collection
        • Build the catalog
        • Operations
        • Use a collection
        • Dashboard
        • Recent activity
      • Document intelligence
        • The visual workbench
        • Plain English rules
        • Audit-grade citations
      • Skills
      • Verified actions
      • Build patterns
      • FAQ
    • Use
      • Home and ad hoc work
      • Sai
      • Chat with a project
      • Work items
        • Triage your inbox
        • Review AI output
        • Correct and send back
      • Tasks
      • Insights
      • Memories you own
      • Talking to your agent well
      • FAQ
    • Observe and measure
      • Insights
      • Analytics
      • Audit deep dives
      • Notifications and alerts
    • Develop
      • Custom actions
      • Authoring MCP servers
      • Skills SDK
      • Agent SDK
      • Local development environment
    • DEPLOY
    • Self-host Sema4.ai
      • Requirements
      • Cluster and node requirements
      • Install the sandbox runtime
      • Deployment targets
      • AWS EKS
      • Reference
      • Identity providers (OIDC)
        • Microsoft Entra ID
        • Auth0
        • Okta
      • Network endpoints
    • REFERENCE
    • API reference
      • Authentication
      • Projects API
      • Runs API
      • Tasks API
      • Work items API
      • Integrations API
      • Audit API
    • Solutions
      • AP help desk
      • Payment remittance
      • Invoice exception management
    • SUPPORT
    • Support
      • Troubleshooting
      • Terminology
      • Browser support and known issues
      • Contact us
Build agents
Develop AI actions
Actions with data

Create actions with data

Data access is a powerful feature in Sema4.ai Studio that allows you to build agents that can access enterprise data and make predictions based on machine learning models.

Visit the solution tutorial Build agents with Data Access to learn how to build agents with data access.

Create custom actionPublish to Studio

  • Product
  • Use Cases
  • Blog
  • Events
  • Studio

© 2025 Sema4.ai All rights reserved.

  • Terms and Conditions
  • Privacy Policy
  • Cookie Policy