How it works
This page is under construction.
What are the recommendations based on?
rek.ai’s recommendations are based on an AI model trained on anonymous interaction data collected from visitors’ behavior on the website.
The system gathers signals such as:
- which pages are visited (page names, IDs, URLs, referrers, recently visited pages)
- visit patterns (views today, last 30 days, days since last visit, session start, browsing history)
- browser/device context (mobile or not, browser language, local time)
- metadata from the pages (titles, descriptions, images, JSON-LD fields like cost, department, etc.)
- optional custom features provided by the website owner (e.g. department on an intranet).
This data is then stored on rek.ai’s servers (without saving IP addresses or personal data). Once enough interactions are collected, a deep learning model is trained (typically daily) to predict what content will be most relevant for each visitor.
In practice, this means that rek.ai’s recommendations are personalized suggestions generated by an AI model, trained on behavioral signals, content metadata, and optional custom features—always based on anonymous usage patterns rather than personal identifiers.
Data gathering
Data processing
Show Recommendations
Implementation examples
Get inspired by our customer cases and learn how to use Rek.ai to improve your business.
