XSearch Overview

Prev Next

XSearch Overview

XSearch is XGen AI’s revolutionary generative AI product that generates search results based on customer/shopper prompts in a search bar on a website, rather than keywords indexed in the product catalogue. When shoppers engage with XGen’s generative AI search, they can look for:

  • Items for a specific occasion (weddings, galas, birthday parties, business events, etc.)

  • A combination of colors (e.g. black and white shoes) or color attributes (dark, light, bright, etc.)

  • Pop culture aesthetics (I want to look like Barbie)

  • Location specific looks (dark London vibes, or Miami Beach vibes, etc.)

See example below

Valentino's online shop displays various blue dresses for wedding occasions.

Why We Built This

Existing solutions are ill equipped to handle the complex ways in which shoppers search for products online, often producing ‘0 results found’ while also providing irrelevant alternatives to display something rather than leaving the area empty. The inspiration behind XGen’s generative AI models was to eliminate ‘0 results found’ once and for all, and to provide shoppers with highly relevant results, regardless of how they’ve asked it.

How is Generative Search Different?

Generative search is built on a collection of high-performance models developed by OpenAI, Anthropic and other leading AI companies. As a result, it offers you 100x the intelligence, requires less maintenance effort, and delivers a far superior performance.

Legacy AI

Generative AI

Data Dependence: Relies solely on your existing data, meaning poor data quality leads to poor results. Garbage in, garbage out.

Real-Time Data: Sophisticated models overcome data limitations in real-time, eliminating product feed enhancement.

Segmented Recommendations: Generic recommendations tailored to large groups, missing the mark on individual preferences

Hyper-Personalization: Real-time, predictive AI generates a unique product feed for every user, adapting to their behavior at every interaction.

Cookie-Cutter Models: Uses the same black-box models as your competitors, offering no competitive advantage.

Brand-Aligned Models: AI models designed and trained specifically for your shoppers and your brand.

Slow & Inflexible: Slow iteration cycles and ill-suited for seasonal promotions, special collections, or rapidly changing trends.

Agile: Allows you to iterate faster and well suited for seasonal promotions or special collections

Features and Functionality

In this guide, we will walk you through how to configure Search to best meet your business needs. We will cover:

  • Getting started and installation guidelines

  • An SDK toolkit to share with your IT or technology team to implement on your brand’s website, so you can access a shopper’s search history

  • XSearch Sandbox

  • XSearch Models

  • XSearch Engines

  • XSearch Experiences

  • XSearch Keywords

  • XSearch Analytics

  • Merchandising Rules