How AWS Radar Works
Your personal AWS announcement filter.
Stop drowning in 50+ weekly AWS announcements. AWS Radar uses AI to pre-filter and learns from your feedback to surface only what matters to your stack.
The Problem
AWS publishes 50+ announcements per week. Most aren't relevant to you:
- New region launch in Asia? Irrelevant if you're EU-only.
- GameLift update? Irrelevant if you don't do gaming.
- Yet another SageMaker feature? Irrelevant if you're not doing ML.
Reading everything wastes hours. Missing the important ones costs more.
The Solution
AWS Radar learns what matters to YOU:
- AI Pre-filtering — Nova Lite analyzes each announcement, generates summaries, extracts topics
- Your Feedback — Vote on tags and services that matter (or don't)
- Personalized Feed — System learns, ranks by relevance, surfaces what you care about
Smart Sorting
Every announcement gets a relevance score based on:
- Your tag votes: Upvoted "serverless"? Those rank higher.
- Your service preferences: Marked "GameLift" irrelevant? Filtered out.
- Time decay: Recent announcements rank higher (0.05/day decay)
display_score = base + tag_adjustments - (days_old x 0.05)
Tag Voting
Each announcement has topic tags (e.g., "serverless", "security", "cost-optimization").
- Upvote a tag: all announcements with that tag rank higher
- Downvote: they rank lower
- Vote once, affects all future announcements with that tag
Service Preferences
In Settings, mark any of 256 AWS services as:
- Relevant: always show
- Irrelevant: filter from "Relevant" tab
- Neutral: default, scored normally
Mark "GameLift" irrelevant once. Never see GameLift announcements again.
Emergent Taxonomy
Tags aren't predefined—they emerge from usage:
- AI suggests tags freely
- Similar tags get matched via embeddings
- Tags appearing 4+ times get "promoted" to official taxonomy
- Community usage shapes the vocabulary
Star & Flag
- Star: Bookmark for later (separate tab)
- Flag: Dismiss permanently (filtered from Relevant tab)
FAQ
How long until I see results?
Immediately. Vote on 5-10 announcements and you'll see sorting change. Service preferences take effect on next page load.
Does it learn across users?
Tag taxonomy: Yes—tags promoted by community usage.
Preferences: No—your votes are private to your account.
What about new AWS services?
New services get added automatically when AWS announces them. They start as "neutral" in your preferences.
How is relevance calculated?
base_score = 3 (fixed baseline) for each tag: if upvoted: +1 if downvoted: -1 display_score = base_score + adjustments - (days_old x 0.05)
Is my data private?
Your preferences stay in your account. Announcements are public AWS data. AI processing uses Amazon Bedrock (AWS-hosted, no data leaves AWS).
The Goal
Spend 5 minutes instead of 30 minutes per week on AWS announcements.
Open the Relevant tab. Skim the top 10. Done.