Intelligence · Circo

Personalised feeds on a recommendation engine

Every interaction batched into a recommendation engine, with visibility rules and session continuation so the feed stays safe and attributable.

Architect · Engineering lead

What it does

Views, bookmarks, reactions, comments, shares and series activity are batched through a queue into a recommendation engine. Feeds are requested per user, and follow-up pages continue the same recommendation session, so a view can be attributed to the recommendation that produced it.

The rule that matters

A feed only ever serves items that are published, transcoded and not deleted. That visibility rule is enforced at query time, on every request.

System

Circo Sparks: one event model for engagement

Views, likes, shares, comments and recircos turned into a single event stream, where every action carries value for both the fan and the creator.

EventsAnalyticsRewards

System

Audience affinity: what should I make next?

Recommendation turned around for creators: from your audience, find the topics they watch elsewhere that you haven't made, with privacy floors in code.

AnalyticsPrivacy

Personalised feeds on a recommendation engine

What it does

The rule that matters

Related systems

Circo Sparks: one event model for engagement

Audience affinity: what should I make next?

What it does

The rule that matters