Laminar allows you to send custom events to the platform. This can be useful for tracking any events that are not covered by the spans and tracing.

​
Introduction

You can think of events as a data payload attached to a single timestamp. Main difference between events and spans is that events have a single timestamp and don’t have a duration. Events allow you to track any free-form data that happens in your application You can then run queries on the events to get event-based analytics and insights. You can also build custom dashboards to visualize the events.

​
Sending events

To send an event, simply use the event function on Laminar class. 
import { Laminar } from "@lmnr-ai/lmnr"

Laminar.initialize();

Laminar.event({
    name: "my-event",
    attributes: {
        "user-feedback": "positive",
        "user_data": JSON.stringify({
            "name": "John Doe",
            "age": 30,
            "email": "john.doe@example.com"
        })
    }
})

​
Span context

If you create an event inside a span context, e.g. in an observed function, the event will be associated with the current span. Otherwise, we will create a new span for the event and attach the event to it.

​
Timestamp

By default, event will have the current timestamp. You can also specify the timestamp manually: 
import { Laminar } from "@lmnr-ai/lmnr"

Laminar.initialize();

Laminar.event({
    name: "my-event",
    timestamp: new Date("2025-01-01T00:00:00.000Z")
})

​
Attributes

Attributes are passed as an object/dictionary. Allowed attribute types are string, number, boolean, string[], number[], boolean[].
If you want to pass a JSON object as an attribute, you need to stringify it first.

​
Specifying session or user id

You can associate an event with a session id or user id. 
import { Laminar } from "@lmnr-ai/lmnr"

Laminar.initialize();

Laminar.event({
    name: "my-event",
    sessionId: "123",
    userId: "456"
})

​
Querying events

You can query events using the events table in Laminar SQL Editor. 
SELECT * FROM events
WHERE session_id = '123'
AND timestamp >= '2025-01-01T00:00:00'