Monetize360

Using Agents

Learn how to interact with AI agents through the chat interface

Using Agents

Agents provide a conversational interface for interacting with your data and performing tasks. This guide explains how to use agents effectively.

Accessing Agents

From Agents Page

  1. Navigate to the Agents page
  2. Find the agent in the list
  3. Click on the agent name or "Chat" button
  4. The chat interface opens

From Other Pages

Agents can be accessed from:

  • Builder Pages: Agents linked to Builder pages
  • Dashboard Widgets: Agents embedded in dashboards
  • Direct Links: Direct links to agent chat interfaces

Chat Interface

Interface Overview

The chat interface consists of:

  • Agent Header: Agent name and description
  • Chat Messages: Conversation history
  • Input Area: Text input for queries
  • Send Button: Submit query to agent

Starting a Conversation

  1. Enter Query: Type your question or request in the input area
  2. Submit: Click "Send" or press Enter
  3. Wait for Response: Agent processes your query and responds
  4. Continue Conversation: Ask follow-up questions or make additional requests

Query Examples

Data Questions:

  • "Show me all customers from last month"
  • "What is the total revenue for Q1?"
  • "List all pending orders"

Task Requests:

  • "Generate a report for customer orders"
  • "Create a workflow for order processing"
  • "Find all customers with overdue invoices"

Analysis Requests:

  • "Analyze sales trends for the last 6 months"
  • "Compare revenue across different regions"
  • "What are the top 10 products by sales?"

Agent Capabilities

Data Queries

Agents can query your data:

  • Object Queries: Query Object records using natural language
  • Filtered Queries: Apply filters and conditions
  • Aggregated Queries: Calculate sums, averages, counts, etc.
  • Joined Queries: Query across multiple Objects

Example: "Show me all orders with status 'pending' from the last 7 days"

Report Generation

Agents can generate reports:

  • Dynamic Reports: Generate reports based on queries
  • Custom Reports: Create custom report configurations
  • Chart Generation: Generate charts and visualizations
  • Data Export: Export report data

Example: "Generate a revenue report for last quarter grouped by product"

Workflow Information

Agents can provide information about workflows:

  • Workflow Definitions: Explain workflow configurations
  • Workflow Usage: Show how workflows are used
  • Workflow Steps: Explain workflow steps and logic

Example: "Explain how the order processing workflow works"

Context-Aware Responses

When vectors are configured:

  • Semantic Understanding: Understand meaning, not just keywords
  • Context Retrieval: Retrieve relevant context from vectors
  • Data-Aware: Responses based on your actual data
  • Domain-Specific: Understand your domain and terminology

Conversation Features

Multi-Turn Conversations

Agents support multi-turn conversations:

  • Context Memory: Agent remembers previous conversation
  • Follow-up Questions: Ask follow-up questions naturally
  • Conversation History: View conversation history
  • Context Continuity: Maintain context across turns

Example:

  • User: "Show me customer orders"
  • Agent: [Shows orders]
  • User: "What about last month?"
  • Agent: [Shows last month's orders]

Clarification

Agents ask clarifying questions when needed:

  • Missing Information: Ask for missing information
  • Ambiguous Queries: Clarify ambiguous requests
  • Confirmation: Confirm understanding before proceeding

Example:

  • User: "Generate a report"
  • Agent: "What type of report would you like? Please specify the data source and time period."

Best Practices

Query Formulation

  • Be Specific: Provide specific details in queries
  • Natural Language: Use natural language, not keywords
  • Include Context: Include relevant context
  • Clear Requests: Make clear, actionable requests

Effective Queries

Good Queries:

  • "Show me all customers who placed orders in the last 30 days"
  • "Generate a revenue report for Q1 2024 grouped by product category"
  • "What are the top 10 products by sales volume this year?"

Less Effective Queries:

  • "customers orders"
  • "report"
  • "data"

Follow-up Questions

  • Build on Previous: Build on previous responses
  • Be Specific: Ask specific follow-up questions
  • Provide Context: Reference previous conversation when needed

Understanding Responses

  • Read Carefully: Read agent responses carefully
  • Verify Data: Verify data in responses when critical
  • Ask for Clarification: Ask for clarification if needed
  • Request Details: Request more details if responses are too brief

Troubleshooting

No Response

If agent doesn't respond:

  • Check Connection: Verify internet connection
  • Wait Longer: Some queries take time to process
  • Try Again: Retry the query
  • Simplify Query: Try a simpler query

Incorrect Responses

If agent provides incorrect responses:

  • Clarify Query: Rephrase or clarify your query
  • Provide More Context: Provide more context
  • Check Agent Configuration: Verify agent has correct vectors and tools
  • Review System Prompt: Check if system prompt needs updates

Tool Errors

If tools fail:

  • Check Permissions: Verify agent has necessary permissions
  • Verify Data: Ensure data exists and is accessible
  • Check Configuration: Verify tool configuration
  • Review Logs: Check error logs for details
Creating AgentsAPI Reference
On this page

On this page

Using Agents
Accessing Agents
From Agents Page
From Other Pages
Chat Interface
Interface Overview
Starting a Conversation
Query Examples
Agent Capabilities
Data Queries
Report Generation
Workflow Information
Context-Aware Responses
Conversation Features
Multi-Turn Conversations
Clarification
Best Practices
Query Formulation
Effective Queries
Follow-up Questions
Understanding Responses
Troubleshooting
No Response
Incorrect Responses
Tool Errors
Related Introduction