HYPERDRIFT
← Back to all posts

Need to Know: AI-Curated News Without the Noise

•
#need-to-know#ai#grok#news#nextjs#supabase#xai

Need to Know: AI-Curated News Without the Noise

A minimalist daily signal feed for focused minds.

The Attention Economy Problem

Traditional news feeds are designed to maximize engagement, not inform you. Endless scrolling, sensational headlines, and algorithmic rabbit holes steal your attention without adding value.

Need to Know takes a different approach: let AI do the filtering so you can focus on what actually matters.

How It Works

Grok-Powered Curation

Each news item is processed by xAI's Grok model (grok-3-mini-beta):

const response = await xai.chat.completions.create({
  model: 'grok-3-mini-beta',
  messages: [{
    role: 'user',
    content: `Analyze this news for professionals interested in ${topics.join(', ')}`
  }]
});

Impact Scoring

Every article gets a 1-10 impact score based on:

  • Market relevance
  • Time sensitivity
  • Action potential

Low-impact noise never reaches your feed.

Premium Metrics

For premium users, deeper analysis includes:

MetricDescription
Market RelevanceHow this affects markets/industry
Security ImplicationsCybersecurity and risk factors
Regulatory ImpactPolicy and compliance changes
Innovation ScoreTechnical advancement significance
Adoption PotentialLikelihood of widespread adoption

Brave Search Integration

Logged-in users get instant Brave Search links for each headline—privacy-respecting deep dives without leaving the ecosystem.

Topic Categories

Free Tier:

  • Crypto
  • DeFi
  • AI

Premium Tier:

  • NFTs
  • Web3
  • DAOs
  • Custom topics

Tech Stack

ComponentTechnology
FrameworkNext.js 15, React 19
AIxAI Grok API
AuthSupabase
CachingRedis (24-hour TTL)
StylingTailwind CSS, shadcn/ui
RuntimeBun

The Freemium Model

Need to Know uses a sustainable freemium approach:

  • Free: 3 topics, cached summaries, basic metrics
  • Pro: All topics, real-time analysis, full metrics, export/save

Cached content is refreshed every 6 hours, balancing freshness with API costs.

Smart Caching

AI API calls are expensive. Our caching strategy:

const cacheKey = `news:${topic}:${date}`;
const cached = await redis.get(cacheKey);

if (cached && !isStale(cached)) {
  return JSON.parse(cached);
}

const fresh = await grokAnalyze(topic);
await redis.setex(cacheKey, 86400, JSON.stringify(fresh));
return fresh;

This keeps the service affordable while ensuring timely updates.

Philosophy: Less is More

Need to Know is intentionally minimal:

  • No infinite scroll
  • No engagement metrics visible to users
  • No algorithmic recommendations beyond topic selection
  • Clean, focused interface

The goal isn't to keep you on the platform—it's to inform you and get out of the way.

What's Next

  • Custom Topic Training: Train the AI on your specific interests
  • Email Digests: Morning briefings delivered to your inbox
  • API Access: Build your own integrations

Try it: needtoknow.hyperdrift.io

Stay informed without losing your focus.

← All posts2026