Feature
Personalized
Suggestions
AI-powered recommendations based on your preferences and engagement. Bloom.ai learns what matters to you and surfaces the most relevant events — automatically.
From raw interactions
to personalized intelligence
Select preferences
Choose up to 5 of 17 categories during onboarding — from Economy and Conflict to Science and Health.
Interact naturally
Click, read, copy, and share events. Every interaction is tracked and weighted to build your profile.
Build your vector
Bloom.ai computes a 768-dimensional interest vector in the same embedding space as events.
Get ranked results
Events are scored using a hybrid formula and ranked per-user at read time for maximum relevance.
Scoring engine
Hybrid actionability scoring
Every event is scored against your unique profile using four weighted signals. The result is a feed that adapts to your evolving interests without manual configuration.
User Vector
768-dim embedding
Event Vector
768-dim embedding
Actionability Score
50%
Similarity
25%
Severity
15%
Recency
10%
Specificity
Your personalized feed
Deduplicated, ranked, and refreshed in real time
Capabilities
Everything that powers the personalized suggestions engine.
Category preferences
Users select up to 5 of 17 categories — from Economy to Conflict to Science — used to filter and prioritize the dashboard feed.
Interest vectors
User engagement is tracked and converted into a 768-dimensional interest vector in the same embedding space as events, enabling semantic matching.
Similarity scoring
Cosine similarity between user interest vectors and event vectors produces a personalized relevance score for every event.
Hybrid actionability
Final ranking combines vector similarity (50%), event severity (25%), recency (15%), and specificity (10%) into a single actionability score.
Smart notifications
A scheduled job evaluates the top matching events and sends personalized suggestion notifications once every 24 hours.
Adaptive ranking
As users interact with events, their interest vector evolves — recommendations improve over time without manual configuration.
Engagement signals
Every interaction counts
Bloom.ai tracks implicit signals from your natural usage patterns — clicks, reading time, sharing, and more. Each signal is weighted to capture the depth of your interest, not just surface engagement.
- High-weight signals: screenshots, sharing, text copy
- Medium-weight: dwell time and deep reading
- Negative signals: quick bounces reduce noise
Your feed, your focus
Start using Bloom.ai and let the personalization engine learn what matters to you.