Snowplow Assistant
The Snowplow Assistant is an AI assistant embedded in Snowplow Console. It lets you manage your tracking implementation, monitor pipelines, troubleshoot issues, and configure Signals through natural language conversation.
How it works
The Snowplow Assistant runs as a chat interface inside the Console. When you send a message, the assistant determines which tools and skills to use, calls the relevant Snowplow APIs on your behalf, and returns the results in a conversational format.
Permissions
The Snowplow Assistant operates with your existing Console permissions. It authenticates using your current session and can only perform actions that your account is authorized to do. It cannot escalate privileges or access resources outside your organization.
Any action that modifies data (creating a schema, updating an enrichment, deleting an alert) requires your explicit confirmation before the assistant proceeds.
Data privacy
The Snowplow Assistant does not have access to personal data or PII such as names or email addresses. It operates only on Snowplow configuration and metadata (schemas, tracking plans, pipeline settings, failed event error details, and similar).
To provide this functionality, the content of your messages and the configuration data the assistant retrieves on your behalf are sent to a third-party LLM provider for processing. Do not enter personal data, credentials, or other sensitive information into the chat.
Enabling and disabling the assistant
You can enable or disable the Snowplow Assistant for your organization from the Settings section of the Snowplow Console. When disabled, the chat interface does not appear in the Console for any user in your organization.
When enabling the feature for the first time, an admin user must accept the terms and conditions related to LLM usage within the Snowplow product.
Usage Limits
The Snowplow Assistant has weekly refreshing token limits at the organisation level. The limit is set at a level that under normal usage you will not reach it.
Capabilities
The assistant provides support across several areas of the Console. It loads specialized skills on demand based on your request.
Tracking design
Design and implement your event tracking using natural language:
- Create and manage data structures (Iglu schemas)
- Create and manage event specifications with entity contexts
- Create and manage tracking plans
- Search Iglu Central for reusable public schemas
- Browse the data catalog to understand what is already being tracked
Implementation guidance
Get help instrumenting Snowplow trackers in your applications:
- Generate copy-paste-ready tracking code for web, mobile, and server-side platforms
- Look up schema definitions and event specifications for correct implementation
- Follow a guided workflow from instrumentation through to production deployment
Pipeline management
Explore and understand your pipeline infrastructure:
- List pipelines with their status, collector endpoints, and labels
- View real-time pipeline metrics including throughput, enrichment latency, and bad row rates
- Review collector configuration (CNAME, cookie policy, DNS)
- List and manage enrichments
- View mini and Micro development instances
Troubleshooting
Diagnose and resolve pipeline issues:
- Investigate failed events with detailed error breakdowns
- Analyze schema validation failures and enrichment errors
- Drill into specific errors by app ID and tracker version
- Get targeted recommendations for fixes
Data quality
Manage alerts for data quality issues:
- Create and manage data quality alerts via email or Slack
- Filter alerts by app ID, issue type, or data structure
- Update or remove existing alert configurations
Console operations
Perform day-to-day management tasks:
- View and manage source applications, including their platform, app IDs, and entity definitions
- Navigate directly to relevant Console pages from the chat
- Search and browse the Snowplow documentation
Signals
If Signals is enabled for your organization, the assistant can help you configure and manage it:
- Define and manage attribute keys and attribute groups
- Create services for pull-based attribute access
- Set up interventions for push-based actions
- Test attribute groups against sample data
- Publish and unpublish configurations
- Query real-time attributes for specific users