Built for recommendation teams

Deploy Recommendation Instances for Every Store and Channel

Kliki gives teams a clear SaaS workflow for deploying engines, streaming events, and serving personalized recommendations from one platform.

Kliki
Workspaces
Events
API Keys
Insights

Why Teams Use Kliki

Kliki replaces scattered scripts and ad hoc dashboards with one recommendation control plane.

Traditional Setup

  • Shows the same products to every customer
  • Needs weeks of setup and data collection
  • Expensive to maintain and update
  • Gets it wrong for new customers
Kliki
  • Learns from live traffic immediately
  • Updates with every event and refresh
  • Simple setup with minimal overhead
  • Adapts to each customer and catalog

Built for recommendation operations

Kliki keeps catalog, event, and recommendation flows visible so teams can deploy and monitor with confidence.

Retail personalization

Show related products at the moment a shopper is deciding what to add next.

Content discovery

Use recent activity to surface the next article, video, or collection that fits the session.

Marketplace ranking

Bring long-tail items into the recommendation flow instead of letting them disappear in the catalog.

Connect your stack

Use the dashboard to provision a workspace, copy the API key, and connect your storefront or backend.

Workspace isolation: Each customer account keeps its own data and key.
Live instance status: Monitor health, uptime, and event flow from the dashboard.
API-first workflow: Send events and fetch recommendations from the same control plane.
Workspace
Instance
API
Insights

Deploy, Observe, and Iterate.

  • Clear setup: Create the workspace, instance, and key in one flow.
  • Always current: Events and live status update as traffic arrives.
  • API-first: Connect storefronts, marketplaces, or content apps.
See the docs
recommender.ts
// 1. Initialize your client with the workspace API key
const engine = new Kliki(API_KEY);

// 2. Track the user action for this instance
await engine.track({
  user: 'user_99',
  action: 'purchase',
  items: ['prod_1', 'prod_2'],
  metadata: { wilaya: 'alger', platform: 'mobile' }
});

// 3. Request recommendations for the same instance
const suggestions = await engine.getSimilar(
  'user_99', 
  { limit: 10 }
);

Ready to launch recommendations?

Create a workspace, connect events, and start serving recommendations.