Custom Agents

Build AI agents tailored to your organization’s specific needs.

Overview

Custom agents allow you to create AI-powered assistants with:

• Custom instructions — Define how the agent responds
• Enterprise knowledge — Connect internal data sources
• Tool integrations — Access external systems and APIs
• Configurable behavior — Control tone, scope, and capabilities

Agent Types

Prompt Agent

LLM-powered agent with custom instructions.

Type: Prompt Agent
Use Case: General Q&A, content generation, analysis

Configuration:
  - System prompt with instructions
  - Model selection (GPT-4, Claude, etc.)
  - Temperature and output settings

Best for: Flexible conversations, content creation, general assistance

API Agent

Connect to external APIs and services.

Type: API Agent
Use Case: Data retrieval, system integration, actions

Configuration:
  - API endpoint configuration
  - Authentication (API key, OAuth)
  - Request/response mapping

Best for: Real-time data access, triggering external actions

Search Agent

RAG-based retrieval from connected knowledge sources.

Type: Search Agent
Use Case: Enterprise search, documentation Q&A

Configuration:
  - Knowledge source connections
  - Retrieval settings
  - Answer generation model

Best for: Finding information in documents, knowledge bases

Agentic Flow Agent

Multi-step workflows with decision logic.

Type: Agentic Flow Agent
Use Case: Complex processes, approvals, multi-step tasks

Configuration:
  - Flow definition with nodes
  - Conditional branching
  - Human-in-the-loop steps

Best for: Guided processes, multi-step automation

Autonomous Agent

Self-directed agent that completes tasks independently.

Type: Autonomous Agent
Use Case: Research, analysis, complex automation

Configuration:
  - Goal definition
  - Available tools
  - Guardrails and limits

Best for: Open-ended tasks, research, comprehensive analysis

Bot Agent

Conversational dialog with structured flows.

Type: Bot Agent
Use Case: Guided interactions, forms, structured data collection

Configuration:
  - Dialog flows
  - Entity extraction
  - Fallback handling

Best for: Form filling, guided workflows, structured conversations

Workflow Agent

Visual workflow automation.

Type: Workflow Agent
Use Case: Business process automation

Configuration:
  - Visual workflow builder
  - Integration nodes
  - Trigger conditions

Best for: Automating repetitive business processes

MCP Agent

Model Context Protocol integrations for external tools.

Type: MCP Agent
Use Case: External tool connections, third-party integrations

Configuration:
  - MCP server connections
  - Tool selection
  - Permission scoping

Best for: Connecting to external systems via MCP protocol.

Create a Custom Agent

Step 1: Start Creation

1. Navigate to Custom Agents
2. Click Create Agent
3. Select the agent type

Step 2: Configure Basics

Name: "Engineering Assistant"
Description: "Helps engineers find documentation and answer technical questions"
Icon: Select or upload

Step 3: Set Instructions

Write clear instructions that define agent behavior:

You are an engineering assistant for the platform team.

Your responsibilities:
- Answer questions about our codebase and architecture
- Help find relevant documentation
- Explain technical concepts clearly

Guidelines:
- Be concise and technical
- Include code examples when helpful
- Cite documentation sources
- Escalate unclear issues to the team lead

Step 4: Connect Knowledge

Add enterprise knowledge sources:

Source Type	Examples
Search AI	Confluence, SharePoint, Google Drive
Agentic Apps	Agent Platform applications
Amazon Q	AWS-based knowledge

Configuration:

1. Click Add Knowledge Source
2. Select connector type
3. Authenticate and configure
4. Set retrieval parameters

Step 5: Add Tools (Optional)

Connect external tools and APIs:

Tools:
  - jira_integration:
      description: "Create and search Jira tickets"
  - github_api:
      description: "Search code and create issues"
  - calendar_tool:
      description: "Check availability and schedule meetings"

Step 6: Test and Deploy

1. Use the Preview panel to test conversations
2. Iterate on instructions and configuration
3. Deploy to selected channels
4. Monitor performance in Analytics

Enterprise Knowledge

Search AI Connector

Connect to Search AI for RAG-based retrieval:

1. Navigate to Enterprise Knowledge → Search AI
2. Select your Search AI application
3. Configure retrieval settings:
   • Top K results — Number of chunks to retrieve
   • Similarity threshold — Minimum relevance score
   • Source filtering — Limit to specific content sources

Agentic Apps Connector

Use Agent Platform apps as knowledge sources:

1. Navigate to Enterprise Knowledge → Agentic Apps
2. Select the agentic app to connect
3. Configure access permissions

Amazon Q Connector

Connect Amazon Q for AWS-based knowledge:

1. Navigate to Enterprise Knowledge → Amazon Q
2. Configure AWS credentials
3. Select Amazon Q application
4. Set access permissions

Agent Management

Versioning

Track changes to agent configurations:

• View version history
• Compare versions
• Roll back to previous versions
• Tag versions for releases

Access Control

Configure who can access and modify agents:

Role	Permissions
Owner	Full control, delete
Editor	Modify configuration
Viewer	View only
User	Interact with deployed agent

Monitoring

Track agent performance:

• Conversation volume
• Response quality metrics
• Error rates
• User satisfaction

Best Practices

Instructions

• Be specific about agent scope and limitations
• Include examples of good responses
• Define escalation paths for edge cases
• Specify tone and communication style

Knowledge

• Keep knowledge sources up to date
• Use specific source filtering for accuracy
• Test retrieval quality regularly
• Monitor for outdated information

Testing

• Test edge cases and error scenarios
• Validate tool integrations
• Check guardrail effectiveness
• Get feedback from target users

Related

• Pre-built Agents
• Pre-built Solutions
• Analytics