i6RecSys

Decision family: Recommendation, Personalization, Propension & Assortment

Models individual and collective behavior from browsing, purchase and interaction events to rank what to offer, to whom and at what moment. Covers everything from in-product recommendation to assortment design per store and prioritization of contact lists.

Output types

• Ranked list of items per customer or session.
• Propensity score for purchase, repurchase and churn.
• Prioritized audience per campaign and channel.
• Suggested assortment per store, region or cluster.
• Cross-sell and up-sell combinations.
• Drivers of each recommendation (affinity, context, history).

Engines involved

i6RecSys and i6Previsio.

How to call

await sdk.recsys().recommend({ dataset, table, context })

Returns: a Promise with an ordered list of recommended items (id, score).

Minimal example

<script src="https://cdn.infinity6.ai/releases/sdkweb/latest/bundle/i6sdk-web.js"></script>
<script type="module">
  async function main() {
    const sdk = new window.I6Sdk();
    const user = await sdk.auth().requireUser();
    if (!user) return;

    const result = await sdk.recsys().recommend({
      dataset: "<your-dataset>",
      table: "<your-table>",
      context: { userId: "<id>" },
    });
    console.log(result);
  }
  main();
</script>