AI Chat: Advanced Usage Tips and Tricks

To receive the most accurate and useful responses from the AI Chat Agent, apply the following practices:

• Be Specific:
  The more specific the query, the more precise the response. For example, rather than asking “Show me loads,” try “Show import loads for Customer X scheduled for delivery tomorrow.”
• Use Filters:
  Incorporate filters to narrow results. Common filters include status, date range, customer, load type, and location.
• Combine Multiple Criteria:
  Multiple filters can be combined in one query. For example: “Show open invoices over $5,000 for Customer Y that are past due.”
• Use Follow-Up Questions:
  When the initial response is insufficient, follow-up questions can be asked. The agent maintains context during a conversation.
• Clarify Ambiguities:
  If clarification is requested, respond with the necessary details to help the agent better interpret the query.

Additional Tips with New Features

• Organize by Conversation:
  Start a new conversation when changing topics to maintain organized and searchable chat history.
• Use Search Effectively:
  Search chat history using precise keywords. For example, “ABC Corp invoice” will yield more accurate results than simply “invoice.”
• Request Calculations:
  Specific calculations can be requested. Example queries include:

             “Calculate total revenue from Customer X this month”

             “Show average delivery time for loads last week”

• Combine Export Options:
  Data can be both downloaded and emailed. First download for internal use, then email to external stakeholders if needed.

Handling Complex Scenarios

The AI Chat Agent can support complex workflows. Here’s how to manage them effectively:

• Multi-Step Queries:
  Break large requests into smaller ones. For example:

              “Show all loads for Customer X”

              “Now show delayed loads”

              “Filter for LFD within 48 hours”

• Contextual References:
  Follow-up questions can reference previously mentioned data. For example, after asking about a load, a follow-up can be: “Who is the driver for that load?”
• Comparative Queries:
  Ask comparative questions like:

            “Which customer has the most active loads?”

            “Show loads taking longer than average to deliver”

Enhanced Scenario Management

• Multi-Step Analysis:
  Use code execution for layered analysis. Start by retrieving a dataset, then run calculations based on it.
• Data Sharing Workflows:
  Create streamlined workflows combining data retrieval, analysis, and sharing—for example: generate a report, download it, and email it.
• Historical Analysis:
  Leverage past queries for comparisons. For instance:

1. “Show today’s active loads”

2. Later: “Compare this with last week’s active load count”

Troubleshooting Common Issues

When the AI Chat Agent produces unexpected results, consider the following:

• Rephrase the Query:
  If the output is off, try restating the question in simpler terms.
• Check for Typos:
  Verify that identifiers such as load or container numbers are entered correctly.
• Provide More Context:
  If the answer seems unrelated, include additional details.
• Start a Fresh Conversation:
  Begin a new chat if the current one becomes confusing or inconsistent.
• Check System Status:
  If the agent is unresponsive or slow, consult a system administrator or support team.

Advanced Troubleshooting

• Token Errors with External Services:
  Issues with integrations (e.g., Google Sheets or SharePoint) may be related to expired tokens. Wait a few minutes for auto-refresh and try again.
• Download Issues:
  If downloading fails, confirm browser permissions and settings for file downloads.
• Email Delivery Problems:
  Ensure recipient email addresses are valid. If delivery fails, check the spam folder or wait in case of service delays.
• Search Limitations:
  When expected results aren’t found, adjust search terms or try more specific keywords.
• Code Execution Errors:
  Simplify or break down complex calculations to improve success rates.

Keyboard Shortcuts and Efficiency Tips

Use these interface shortcuts and tips for faster, smoother interaction:

• Enter — Send message

• Up Arrow — Edit the most recent message (before response is received)

• Ctrl+F / Cmd+F — Search within a conversation

• Tab / Shift+Tab — Navigate between interface elements

• Escape — Cancel actions or close dialogs

Save Frequently Used Queries:
Maintain a list of commonly used queries for quick reuse.

Use Consistent Terminology:
Standardize terms when interacting with the agent. For instance, always use “loads” instead of alternating between “shipments,” “deliveries,” or other variants.