Skip to main content

AI Agents Overview

AI Agents are specialized intelligent assistants configured with specific knowledge, capabilities, and behaviors to help you with particular tasks and workflows. Unlike general chat, agents come pre-loaded with domain expertise, connected data sources, and custom instructions tailored for specific purposes.

What Makes Agents Different?​

While Galene Chat offers flexible, on-demand AI assistance, agents provide:

  • Persistent Configuration: Pre-defined behavior and knowledge that's always available
  • Specialized Knowledge: Connected to specific files, databases, websites, and external systems
  • Consistent Behavior: Same expert assistance every time you use them
  • Team Collaboration: Shareable with colleagues for consistent workflows
  • Purpose-Built: Designed for specific tasks, domains, or departments

Think of it this way:

  • Galene Chat = General-purpose AI assistant you configure per conversation
  • AI Agents = Specialized experts with permanent knowledge and capabilities

Personal vs Organization Agents​

Personal Agents​

Agents you create and control, private by default:

  • Full Control: Configure exactly how you want
  • Private by Default: Only you can access unless you share
  • Flexible Sharing: Share with specific colleagues who has the required permissions
  • Your Workspace: Create as many as you need

Organization Agents​

Agents created by administrators, available to all or specific groups:

  • Flexible Access: Can be company-wide or restricted to specific roles
  • Centrally Managed: Maintained by administrators
  • Standard Workflows: Ensure company-wide consistency

For information about creating and managing organization-wide agents, see the Admin Documentation.

What Can Agents Do?​

Agents can be configured with various capabilities and connected to different knowledge sources:

Core Capabilities​

Web Search: Access real-time information from the internet

COMING SOON

Code Execution: Run Python code for data analysis
Data Visualization: Generate charts and graphs

Knowledge Integration​

Static Knowledge Base - Fixed content captured at the time of upload:

  • Files and Documents: Upload PDFs, spreadsheets, presentations
  • Websites: Scrape content from web pages

Dynamic Knowledge Base - Live connections that stay synchronized:

  • Cloud Storage: Connect to SharePoint, Google Drive and S3
  • Databases: Query organizational data with natural language
  • External Systems: Integrate via MCP servers (APIs, business tools, and more)
Administrator-Managed Connectors

Knowledge bases, database connectors, and MCP servers are created and configured by administrators. Users with appropriate permissions can view and use these connectors in their personal agents, but cannot create or modify them.

Learn more in Knowledge Sources and Connectors.

Your Agent Journey​

This section will guide you through everything about AI Agents:

  1. Creating Your First Agent
    Learn how to build an agent from scratch or use templates. Covers basic configuration, naming, instructions, and initial setup.

  2. Agent Configuration
    Advanced customization: model selection, temperature settings, system prompts, and capability fine-tuning.

  3. Adding Knowledge Sources
    Upload files, scrape websites, and give your agent access to specific information. Understand how RAG (Retrieval Augmented Generation) works.

  4. Connecting Enterprise Data
    Integrate databases, external APIs, and business systems. Enable agents to query organizational data and interact with external tools.

  5. Agent Examples
    Agents excel across various business functions. Explore some examples.

  6. Sharing Agents
    Collaborate with colleagues by sharing agents. Control access and work together on common workflows.

Getting Started​

Ready to create your first agent?

New to Agents? Start with Creating Your First Agent for a step-by-step walkthrough.

Know What You Need? Jump directly to:

This documentation covers personal agent creation and management. For organization-wide agent administration, see the Admin Agents Documentation.