We build software, dev stacks & apps — and we still offer hosting.
hTech

AI Logs & Event Tracking in the hTech AI Ticket Assistant Print

  • 0

Overview

The hTech AI Ticket Assistant includes a powerful logging system that records every action the AI performs. These logs help you understand how the AI is interacting with tickets, diagnose issues, and monitor automation behavior across your support environment. This guide explains what each log entry means and how to use event tracking effectively.


What Are AI Logs?

AI Logs are recorded whenever the assistant performs an action such as:

  • Generating an automated reply
  • Creating an internal suggestion
  • Routing a ticket to another department
  • Assigning tags to the ticket
  • Blocking a reply due to safety or escalation
  • Failing to generate a reply for any reason

Each log entry includes the timestamp, the ticket ID, and a detailed description of what happened.


Where to View AI Logs

You can access AI logs inside the module:

  • Go to WHMCS admin
  • Select Addons
  • Click "hTech AI Ticket Assistant"
  • Open the "Logs" or "AI Log History" section

Types of Logged Events

1. Reply Events

These logs show when the AI generates a response. Examples:

  • "AI reply generated successfully"
  • "AI reply delayed and added to queue"
  • "AI reply posted to ticket"

2. Suggestion Events

Occurs in Suggest mode, where the AI prepares internal notes for staff:

  • "AI suggestion created for staff review"
  • "Customer reply flagged for manual review"

3. Routing Events

Shows when the AI moves a ticket to another department:

  • "Ticket routed to Billing (tag: refund)"
  • "Auto-routing triggered by tag: login issue"

4. Tagging Events

Records when the AI assigns tags:

  • "Tag added: billing"
  • "Tags detected: setup, firestick"

5. Safety & Escalation Events

Logs when AI determines a reply should not be sent:

  • "Reply blocked due to escalation risk"
  • "Sentiment too negative to auto-reply"
  • "Auto-reply disabled for this message"

6. Error Events

Any failure or error is logged, including:

  • "AI model error: input too long"
  • "API failure – retry scheduled"
  • "Queue delivery failed: ticket closed"

These entries are key for troubleshooting automation issues.


Why Event Tracking Matters

The logging system helps you understand:

  • Why the AI did or did not reply
  • How tags and routing decisions were made
  • Whether escalations or sentiment blocked automation
  • When errors occurred during processing
  • Whether delay timers or queue processing are working

Logs provide transparency and full visibility into AI behavior.


Log Details Explained

Ticket ID

The log entry identifies which ticket the event relates to. Useful for debugging specific customer interactions.

Action Description

A clear summary of what the AI attempted or completed.

Timestamp

Shows when the event occurred, useful for tracking timing issues, cron delays, or repeated events.


Common Log Scenarios

AI Did Not Reply

Check logs for messages such as:

  • "Auto reply blocked for safety"
  • "Ticket flagged – skipping AI processing"
  • "Delay enabled – reply added to queue"

AI Replied Incorrectly

Look for:

  • "Failed to load SOP"
  • "KB lookup returned no matches"

AI Routing Did Not Trigger

Review:

  • "No matching routing rule found"
  • "Tags detected: none"

Queue Issues

Look for:

  • "Queue entry locked – retry delay"
  • "Reply delivered successfully"
  • "Reply delivery failed – ticket closed"

Purging Logs

You can remove logs anytime from the dashboard.

  • Delete – Remove one log entry.
  • Purge – Clear all logs.

Purging logs does not affect ticket data or AI functionality.


Best Practices

  • Check logs when the AI behaves unexpectedly.
  • Use logs to fine-tune department settings or routing rules.
  • Monitor error trends to detect configuration issues.
  • Review safety blocks to improve SOPs or KB clarity.

Summary

The AI Logs system provides clear visibility into the assistant’s behavior. By understanding each event and reviewing logs regularly, you can ensure safe, predictable, and efficient automation across your entire support workflow.

If you need help diagnosing a specific log entry, feel free to open a support ticket with the ticket ID and log details.


Was this answer helpful?

Related Articles

« Back