Rek.ai on an intranet

Rek.ai on an intranet

OG metadata

The first aspect you should consider is that metadata in the form of images and lead paragraphs is included on the pages. These are what we send along when recommendations are presented. You can enable this by adding them to the page's code the same way as for platforms like Twitter or Facebook, using OG tags.

Here are some useful og:tags to add:

<meta property="og:title" content="My title" />
<meta property="og:description" content="Lorem ipsum." />
<meta property="og:image" content="/image.jpg" />

See also: How to make intranet recommendations contain images and texts (opens in a new tab).

Validation: Select "View page source" and check that the OG tags exist in the code and have the correct values.

What do we know about our user?

To personalize in the best possible way, we want to inform the model about the department each employee works in. We first need to find out what values you have access to in the user object.

See also: What parameters do my user have? (opens in a new tab)

Send user information

Once you know what data you have about the employees, you can submit this information with each page view and recommendation.

See also: How you can add page and user information on an intranet (opens in a new tab).

Validation: Check the Payload of the "view" request sent to Rek.ai. The object should contain an array named cf with the data. Read more about validating CF (opens in a new tab).

Validate access

If you wish to ensure that a user has the right to view the pages during a recommendation, you can check each page's access rights with every recommendation.

See also: Validate user access (opens in a new tab).