What is walkerOS

walkerOS is an open-source event data collection platform you run as code. Capture events from anywhere, transform them with mapping, and route them to any destination, all with built-in consent management.

There are no fragile UI configs and no vendor lock-in, so you can version, test, and trust your tracking.

It's built for analytics engineers, data leads, and developers who own analytics on a custom-built app and want to replace a duct-taped GA4 and server-side tag manager stack with a pipeline they control.

Try it now

Send your first event in five minutes → Quickstart

Then: send it to GA4 with mapping →

Working with an AI assistant?

Build and simulate a flow locally with the walkerOS MCP, no account needed: With an AI assistant.

Why walkerOS?

• Open-Source: MIT licensed and free, with no restrictions
• Self-hostable: Run it on your own infrastructure
• Config-as-Code: Every change goes through version control and PRs
• Composable: Mix sources, destinations, and mappings freely
• Developer-first: TypeScript and declarative tagging, with an optional hosted app and MCP/AI-assisted management

---

How it works

Every walkerOS setup follows one pattern:

• Sources: Where events come from (browser DOM, dataLayer, HTTP endpoints, custom code)
• Transformers: Middleware that validates, enriches, routes, or redacts events before and after the collector
• Collector: The processing engine (consent, enrichment, routing)
• Destinations: Where events go (GA4, your API, BigQuery, any service)
• Stores: Shared key-value storage used by transformers and destinations (caching, file serving, session state)

A "flow" connects sources to destinations through a collector. You can have one flow or many, client-side, server-side, or both working together.

The centerpiece: mapping

Mapping transforms events between stages. It's how walkerOS adapts to any source or destination format.

The same transformation language works at every stage, which is what makes walkerOS composable.

Learn more about mapping →

---

Core concepts

Concept	What It Is	Learn More
Flow	Sources + collector + destinations configuration	Flow configuration
Event model	How events are structured (entity action format)	Event model
Mapping	Event transformation rules	Mapping
Consent	Privacy controls at destination and property level	Consent management
Sources	Event capture from browser, server, or custom code	Sources
Transformers	Middleware for validating, enriching, routing, and redacting events	Transformers
Destinations	Event delivery to analytics, APIs, warehouses	Destinations
Stores	Shared key-value storage for caching, file serving, and session state	Stores

---

Two ways to use walkerOS

Both modes use the same architecture. You pick one at deploy time, after your first event (see operating modes).

Integrated mode

The collector lives inside your application code.

import { startFlow } from '@walkeros/collector';
import { sourceBrowser } from '@walkeros/web-source-browser';

const { elb } = await startFlow({
  sources: { browser: { code: sourceBrowser } },
});

Best for: Building into your app, full TypeScript control, dynamic configuration.

Get started with integrated mode →

---

Bundled mode

The collector is a separate artifact you configure with JSON.

{
  "version": 4,
  "flows": {
    "default": {
      "config": { "platform": "web" },
      "sources": { "browser": { "package": "@walkeros/web-source-browser" } }
    }
  }
}

Then walkeros bundle flow.json builds a standalone file.

Best for: Config-as-code, Docker deployments, separate tracking scripts.

Get started with bundled mode →

---

Quick links

Deploying to production?

• Deploy: self-hosted or managed
• CLI Reference: Build and test commands

Already using walker.js?

• Walker.js Docs: Pre-built convenience bundle

Need help choosing?

• Operating Modes: Detailed comparison