Optimization, exposed as a
getAssignment call.

TypeScript SDK, REST API, sub-100ms edge assignments. Real bandit algorithms running on your funnel — wired in with five lines and a reward.

Start free Sign In

TypeScript

REST API

Edge

Open algorithms

app.ts

import { BanditClient } from '@runbandit/sdk';

const bandit = new BanditClient({
  apiUrl: 'https://runbandit.com',
  apiKey: 'your-api-key'
});

// Get the best variant for this user
const assignment = await bandit.getAssignment('headline-test', 'user-123');

// Show the variant content
document.querySelector('#headline').textContent = assignment.config.content;

// Track when they convert
bandit.trackNamedEvent({ name: 'purchase', value: 29.99 });

Full integration, with context

Pass user context to opt into the contextual bandit — device, location, referrer, anything you have. Per-context assignments are a power-user toggle, not the default.

app.ts

import { BanditClient } from '@runbandit/sdk';

const bandit = new BanditClient({
  apiUrl: 'https://runbandit.com',
  apiKey: 'your-api-key'
});

// Get assignment with context for smarter targeting
const assignment = await bandit.getAssignment(
  'headline-experiment',
  'user-123',
  {
    deviceType: 'mobile',
    location: 'US',
    referrer: 'google'
  }
);

// Show the winning variant
document.getElementById('hero').textContent = assignment.config.content;

// Track conversion with value
bandit.trackNamedEvent({
  name: 'purchase',
  value: 29.99,
  metadata: { product: 'premium-plan' }
});

Architecture

From SDK call to optimized assignment in under 100ms.

Your App

SDK Client

Edge API

Cloudflare Workers

Algorithm Engine

In-Memory State

Dashboard

Real-Time Analytics

Convex•Edge Compute•Sub-100ms E2E

Built for production traffic

Real bandit algorithms running at the edge. Honest math, not branded magic.

Self-tuning

Zero-config

Traffic flows toward winning variants automatically — exploration vs exploitation handled by the algorithm.

Posterior-driven

Bayesian

Thompson Sampling out of the box. Confidence is a real number, not a vibe.

Sample-efficient

Low-traffic

Converges on hundreds of users per day, not the tens of thousands a fixed split needs.

Context-aware

Contextual

Pass device, location, or custom features and opt into per-context assignments via the contextual bandit.

Restless-ready

Non-stationary

Discounted or sliding-window memory, so a bandit that has already made up its mind can be talked out of it.

Watch it converge

A live simulation of traffic flowing across three variants as rewards come in. In production, this happens on every request — sub-100ms, at the edge.

Live ExperimentAuto-optimizing

0h elapsed

Variant A

2.1% CVR32%

Variant B

2.8% CVR37%

Variant C

1.9% CVR31%

Traffic automatically shifts to the best-performing variant. No manual intervention needed.

Five steps. Then hands-off.

Drop-in is mostly drop-in — but you do have to wire the reward. We're honest about the work.

Install the SDK

npm install @runbandit/sdk

Initialize the client

new BanditClient({ apiUrl, apiKey })

Get an assignment

bandit.getAssignment(expId, userId)

Wire your reward

bandit.trackNamedEvent({ name, value })

Watch the math run

open dashboard → watch it converge

Grab your API key.

Drop in the SDK, wire your reward, let the math run. Free tier available, no credit card required.

Start free

Optimization, exposed as a getAssignment call.