Automation logs
Automation Logs allow you to capture structured data from your AI agent's automations and turn it into actionable insights.
This feature is designed for agents that use connected tools, API integrations, or MCP integrations. If your agent primarily answers questions without interacting with external systems, we recommend using Topics instead to analyze conversations and identify improvement opportunities.
When to use automation logs
Automation Logs are useful when your AI agent performs actions such as:
Creating support tickets
Scheduling appointments
Updating CRM records
Retrieving ERP data
Calling APIs
Triggering backend processes
Executing MCP tools
If your goal is to understand what users are discussing with your agent, use Topics instead.
Setting up automation logs
To configure Automation Logs:
Open the agent you want to configure.
Navigate to Activity.
Open Automation Logs.
Create a new Automation Log.
Add the columns you want to collect.
Once an Automation Log has been configured, the next automation execution that provides data for the configured fields will automatically create a new log entry.
How automation logs work
Every automation log entry is connected to a conversation.
When an automation runs during a conversation, it can write data to an Automation Log. Each log entry represents a single row of data and always includes:
ID
Date
Conversation reference
Because automations are executed within the context of a conversation, all logged data remains linked to the conversation where the automation was triggered.
Custom columns
In addition to the default fields, you can define your own columns to store business-specific data.
Examples include:
Customer Name
Order Number
Ticket ID
Product Type
Lead Source
Subscription Plan
Revenue
Status
Each Automation Log can contain up to 50 custom columns.
The values for these columns are automatically populated by your automations.
Example
A support ticket automation might create a log entry containing:
A CRM automation could store:
This creates a structured record of the work performed by your AI agent.
Analyzing your agent
Automation Logs provide detailed insight into how your agent interacts with external systems and business processes.
You can use the collected data to:
Track automation outcomes
Monitor API and tool usage
Measure business impact
Identify missing automation opportunities
Discover trends and patterns
Validate automation performance
Because all data is stored in a structured format, it becomes easy to review, filter, and analyze automation activity over time.
Continuous improvement
Automation Logs are particularly valuable for iterative improvement.
By reviewing the data your automations collect, you can perform gap analyses and identify opportunities to improve your AI agent.
For example, you may discover:
Frequently missing information
Automations that are triggered less often than expected
Opportunities for new automations
Missing integrations
Areas where agent instructions can be improved
Regularly reviewing Automation Logs helps ensure your agent continues to become more effective and delivers greater value over time.
Best practices
Only collect data that provides meaningful business insight.
Use clear and consistent column names.
Start with a small set of important fields.
Review collected data regularly.
Use the insights to improve automations and agent behavior.
Use Topics for conversation analysis and Automation Logs for operational and business process analysis.
Automation Logs transform your agent's actions into measurable data, giving you the visibility needed to optimize and scale your AI-powered workflows.