Our AI Personalization Engine tracks browsing behavior in real time, builds intent profiles per visitor, and renders personalized product feeds — "For You," "Also Bought," "Complete the Look," and "Trending" — that adapt with every click. Privacy-first. GDPR and CCPA compliant from day one.
5
SDK Modules
Real-Time
Scoring
Privacy
First
SDK Modules
Each module is a single HTML element. Drop them into any page — your AI-built storefront, an existing website via the Commerce SDK, or a WordPress site via shortcodes.
Personalized product feed built from each visitor's browsing history, purchase patterns, and real-time intent — no two feeds are the same.
Co-purchase intelligence surfaces products that other customers bought alongside the current product. Powered by real order data, not guesswork.
Multi-signal similarity scoring across taxonomy facets, product type, collection overlap, price range, and co-purchase history for true relevance.
AI assembles complementary products from different categories — if a customer views a dining table, surfaces matching chairs, lighting, and tableware.
Real-time momentum scoring based on views, purchases, and add-to-carts over the last 24–72 hours. Surface what's hot right now, not what was hot last month.
How It Learns
The engine tracks six behavioral signals to build a weighted intent vector per visitor. No cookies required — everything runs on first-party session data with explicit consent.
What visitors search for reveals explicit purchase intent
Browsing patterns show category and price preferences
Strong buying signal — weighted heavily in scoring
Aspirational intent for future purchases
Filtering by color, material, or style reveals taste
Past orders define long-term preference profiles
How It Works
Four steps — consent, signal collection, real-time scoring, and instant rendering.
The SDK shows a privacy notice. Tracking only begins after consent. Fully GDPR and CCPA compliant.
Views, searches, cart adds, purchases, and dwell time build an anonymized intent profile per visitor.
Every product is scored against the visitor's intent vector. Collaborative filtering, content similarity, and co-purchase data are combined.
SDK modules render personalized results instantly. Skeleton placeholders prevent layout shift during loading.
Multi-Signal Scoring
Every candidate product is scored against the visitor's intent profile across multiple axes, then ranked for maximum relevance.
Shared color, material, style, and category facets with viewed products
Products frequently bought together by other customers
Same product type (furniture, fashion, CPG) as browsing history
Products that share collections with viewed items
Products within the visitor's observed price range
Merchant Intelligence
Behavioral data doesn't just power recommendations for shoppers — it gives merchants actionable insights to make smarter inventory and marketing decisions.
See what customers are searching for but not finding. Top search terms with match counts, conversion rates, and demand classification — unmet demand, low conversion, high demand, and emerging trends.
Track which product attributes (colors, materials, styles) are gaining traction. Compare this week vs. last to spot shifts in customer taste before competitors do.
AI segments visitors into behavioral cohorts — new explorers, repeat buyers, price-conscious shoppers, brand loyalists — based on observed patterns, not demographics.
The platform automatically alerts you when high-value demand signals emerge — a search term spiking without matching products, or a category with high views but low conversion.
Cold Start Intelligence
New visitors don't start from zero. The engine reads the landing page context — URL path, search query parameters, referrer data, and UTM campaigns — to build an initial intent profile before a single product is viewed.
If a customer visits from Google searching "modern dining table," the engine immediately prioritizes dining furniture in the "For You" feed. As they browse, the profile refines in real time.
For brand-new stores with no behavioral data yet, the engine falls back to trending products by view and purchase velocity — so recommendation sections are never empty.
URL Path
Product slugs and collection paths reveal category intent
Search Params
Query strings like ?q=modern+sofa provide explicit keywords
Referrer Analysis
Google search terms extracted from the referring URL
UTM Campaigns
Campaign tags indicate audience segment and content affinity
Trending Fallback
When no signals exist, surface what's hot in the store right now
Privacy First
Every tracking decision is transparent, every piece of data is erasable, and nothing happens without explicit consent.
Intent profiles use session IDs — no names, emails, or personal data. Profiles are never shared across stores unless the customer explicitly opts in.
The SDK requires explicit consent before any behavioral tracking begins. No tracking scripts load until the visitor agrees.
Customers can delete all their behavioral data with one API call. Every event, every intent profile, every cross-store record — permanently erased.
Behavioral data is automatically anonymized after 12 months and permanently deleted after 24 months. No indefinite data hoarding.
Behavioral tracking is also protected by bot detection, rate limiting (100 events/min per session), and a strict field allowlist — only product IDs, facet IDs, and event types are stored. No IP addresses, no device fingerprints, no personal identifiers.
The Difference
Most "recommendation" systems just show popular products. Ours builds a unique understanding of every visitor.
| Basic Platforms | Lesuto Smart Discovery | |
|---|---|---|
| For You Feed | Popular products for everyone | Unique feed per visitor based on intent |
| Also Bought | Static manual suggestions | Real co-purchase data from order history |
| Cold Start | Empty sections until data builds | URL context + referrer + UTM on first click |
| Privacy | Third-party cookies | First-party, consent-gated, GDPR compliant |
| Merchant Insights | None | Demand signals, trending facets, segments |
| Data Retention | Indefinite | 12-month anonymize, 24-month delete |
| Cross-Store | Not possible | Opt-in cross-store intent (Hub customers) |
| Cost | $200-500/mo add-on | Free — included in platform |
Works Everywhere
The personalization modules work with every deployment method — AI-built storefronts, Commerce SDK embeds, and WordPress shortcodes.
Sections like "for-you" and "complete-the-look" are available in the section catalog. Add them via conversation or the page builder.
AI Website BuilderSingle div elements that render anywhere. Add data-chameleon="for-you" to any page and the SDK handles the rest.
Commerce SDKShortcodes like [chameleon_for_you] and [chameleon_trending] work with server-side rendering and Auto-Theming.
WordPress PluginConnected Intelligence
The Personalization Engine doesn't just serve shoppers — it feeds intelligence into your merchant AI engines.
Smart Product Discovery is free for every merchant. Every visitor gets a unique experience, every merchant gets actionable intelligence — all with privacy built in from the ground up.
Get Started Free