Skip to content

QuickStart Guide

This guide will walk you through how to get started with Sumatra.

1. Sign up for a free account

To get a free sandbox account, sign up here.

2. Instrument your site or application (no code)

You can send event data to Sumatra using Google Tag Manager, add Sumatra as a destination in Segment or Rudderstack, or use one of the SDKs. The remainder of this guide will assume you are using Google Tag Manager.

In your Tag Manager account, navigate to Templates, click "Search Gallery", and type "Sumatra", and select "Sumatra Integration". This will add a Tag Template that will send pageviews and click events to Sumatra.

create-workspace

Important

Be sure to add your private API key, which can be found on the Settings page of your Sumatra account.

Learn more about API integration

3. Custom orchestration layer & decision logic (included in all paid plans)

From here, the experts at Sumatra will prepare custom data schema and low-code logic for decisions that trigger actions in your product (e.g. personalization, recommendations, and fraud prevention).

Once the orchestration layer is ready, you'll have a production-ready API, full visibility into the code, and self-service access to the console.

Tip

If you want to try setting up the orchestration layer yourself with a Sandbox account, use one of the self-service guides below.

4. Trigger the 'decision' from your application

Add a. .enrich snippet to your application where you want to trigger a decision. The snippet will be specific to your application.

Here's an example snippet for a Personalization use case for an application built with JavaScript.

window.sumatra.enrich('personalize').then((enriched) => {
    // set heading to enriched.features.heading
    // OR set image to enriched.features.hero
});

5. Update, iterate, and experiment

You have self-service access to edit the low-code decision logic as often as you need without changing the API integration.

Next Steps

Self-service guides for Personalization, Recommendation, and Fraud Prevention

Use case Level Guide
Personalization Beginner Personalized text or image by referrer
Personalization Intermediate Using behavioral segmentation for better personalization (coming soon!)
Personalization Advanced Intent prediction with custom ML models (coming soon!)
Recommendations Beginner Recommend similar content (coming soon!)
Recommendations Intermediate Calculate engagement scores for user-specific recommendations (coming soon!)
Recommendations Advanced Optimize a feed by mixing models (coming soon!)
Fraud prevention Beginner Add behavioral signals to Strip Radar (coming soon!)
Fraud prevention Intermediate Detecting suspicious patterns (coming soon!)
Fraud prevention Advanced Using custo ML models to identify fake users (coming soon!)

Additional help