News Terminal
How the Y2 News Terminal Works
The Signal Feed Terminal is Y2's real-time intelligence monitoring system, providing live news signals and AI-generated recaps across multiple topics.
Overview
The News Terminal transforms raw signal data from Gloria AI into actionable intelligence, offering two distinct modes:
Live Feed Mode: Real-time signal stream with instant updates
AI Recap Mode: Synthesized intelligence summaries over configurable timeframes
Unlike traditional RSS readers or news aggregators, the Signal Feed uses AI-powered analysis to filter noise and surface meaningful signals across crypto, AI, macro, and technology topics.
Key Features
Multi-Topic Monitoring
Track 15 specialized intelligence feeds simultaneously:
Crypto: Bitcoin, Ethereum, Solana, Ripple, Aptos, Hyperliquid
AI/Tech: AI Agents, Machine Learning, Tech Industry
Finance: Macro Economics, Real World Assets (RWA), Ondo Finance
Emerging: Virtuals Protocol
Dual-Mode Intelligence
Live Feed Mode
Stream of individual signals as they occur
Sub-minute latency from source to display
Chronological ordering (newest first)
Direct from Gloria AI's signal detection engine
AI Recap Mode
Synthesized summaries across multiple timeframes (12h, 24h, 3d, 7d)
Generated on-demand using LLM analysis
Topic clustering and pattern detection
Executive-style intelligence briefs
Persistent Preferences
Your feed selections, view mode, and timeframe preferences sync across devices:
Automatic save on every change
Instant restoration on page load
No manual "save settings" required
How It Works
Signal Collection Architecture
Live Feed Pipeline
1. Signal Detection (Gloria AI)
Monitors social media, news outlets, on-chain data
Pattern recognition for market-moving events
Sentiment analysis and anomaly detection
Real-time filtering for relevance
2. Categorization (Gloria AI)
Each signal tagged with feed category
Timestamp precision to the second
Metadata preservation (source, confidence, etc.)
3. Delivery (Y2 Backend)
Fetch signals for selected topics
Limit: 50 most recent per request
Deduplication across feeds
Sorted by timestamp (descending)
4. Display (Terminal)
Real-time rendering with React state
Automatic time-ago formatting
Color-coded topic badges
Infinite scroll for history
AI Recap Pipeline
1. Signal Aggregation
Collect all signals for topic within timeframe
Group by semantic similarity
Remove duplicates and low-quality signals
2. LLM Analysis
Send aggregated signals to language model
Extract key themes and trends
Identify market-moving events
Generate executive summary
3. Structured Output
BLUF (Bottom Line Up Front) format
Key developments section
Market implications
Forward-looking analysis
4. Caching
Recaps cached for 1-hour TTL
Regenerate on refresh after expiration
Per-topic, per-timeframe cache keys
Using the Terminal
Selecting Feeds
Multiple Selection Strategy:
Focused Monitoring (1-3 feeds)
Deep attention on specific topics
Easier to follow all signals
Best for active traders or specialists
Example: Bitcoin + Ethereum + Macro
Broad Coverage (4-8 feeds)
Balanced intelligence gathering
Spot cross-topic correlations
Good for portfolio managers
Example: All crypto + AI + Tech
Full Spectrum (9+ feeds)
Complete market awareness
Pattern detection across sectors
Best for researchers and analysts
Use AI Recap mode to synthesize
Feed Selection Tips:
Click any feed to toggle on/off
Selected feeds show gradient backgrounds
At least one feed required to load data
Changes auto-save and trigger refresh
Live Feed Mode
Screenshot placeholder: Terminal showing live signal stream
What You See:
Understanding Signals:
Time Precision
"Just now": < 1 minute old
"42m ago": 42 minutes old
"3h ago": 3 hours old
"2d ago": 2 days old
Signal Quality Gloria AI pre-filters signals, so everything in the feed meets minimum quality thresholds:
Verifiable source (not rumors)
Market relevance (impact on prices/sentiment)
Timeliness (recent developments)
Clarity (understandable, actionable)
Volume Expectations:
High-activity feeds (Bitcoin, AI): 20-50 signals/day
Medium-activity feeds (Aptos, RWA): 10-20 signals/day
Low-activity feeds (Virtuals, Ondo): 5-10 signals/day
Refresh Behavior:
Manual refresh: Click refresh button
Fetches latest 50 signals per selected feed
Merges and deduplicates across feeds
Updates "Last updated" timestamp
AI Recap Mode
Screenshot placeholder: Recap cards with gradient headers
Timeframe Selection:
12 Hours
Best for: Day trading, active monitoring
Coverage: Half-day market cycles
Signal count: 10-25 per topic
Refresh: Every 12 hours recommended
24 Hours
Best for: Daily briefs, overnight catch-up
Coverage: Full trading day
Signal count: 20-50 per topic
Refresh: Every morning
3 Days
Best for: Weekly check-ins, trend identification
Coverage: Short-term trends
Signal count: 60-150 per topic
Refresh: Every 2-3 days
7 Days
Best for: Weekly summaries, strategic planning
Coverage: Full week of developments
Signal count: 140-350 per topic
Refresh: Every Monday
Recap Structure:
Generating Recaps:
Select feeds you want recaps for
Switch to "AI Recap" view mode
Choose timeframe (12h, 24h, 3d, 7d)
Click "Refresh" to generate
Generation takes 10-30 seconds per topic
Recap Caching:
First request generates new recap
Subsequent requests serve cached version
Cache expires after 1 hour
Refresh button always regenerates
Understanding Signal Quality
What Makes a Good Signal?
High-Quality Signal Example:
Low-Quality Signal (Filtered Out):
Signal vs Noise
Gloria AI's filtering removes:
Duplicate reports of same event
Unverified rumors and speculation
Low-engagement social media posts
Sponsored content and advertisements
Historical news (>7 days old unless significant)
This filtering means fewer signals, higher quality—you see 50 meaningful signals instead of 500 irrelevant posts.
Advanced Usage Patterns
Cross-Topic Correlation
Monitor related feeds to spot patterns:
Crypto Correlation
AI/Crypto Convergence
Macro/Crypto Relationship
Event-Driven Monitoring
Breaking News Response:
Select relevant feed (e.g., Bitcoin)
Switch to Live Feed mode
Refresh every 2-5 minutes
Watch for signal clusters (multiple signals on same event)
Post-Event Analysis:
Switch to AI Recap mode
Select 24h timeframe
Generate recap for affected topics
Review synthesized impact assessment
Research Workflow
Daily Intelligence Routine:
Weekly Research:
Integration with Y2 Intelligence
Complementary Systems
Signal Feed (Reactive)
Monitors real-time developments
Responds to market events
Short-term tactical intelligence
User-initiated refresh
Y2 Reports (Proactive)
Scheduled automated research
Deep analysis with recursive search
Long-term strategic intelligence
Automatic delivery
Using Both Together
Example Workflow:
Step 1: Discover Topic (Signal Feed)
Step 2: Configure Deep Research (Y2 Profile)
Step 3: Generate Intelligence Report
Step 4: Monitor Developments (Signal Feed)
Performance & Reliability
Response Times
Live Feed Refresh:
API request: 1-2 seconds
Data processing: <1 second
UI rendering: <500ms
Total: 2-4 seconds for 50 signals
AI Recap Generation:
Signal aggregation: 1-2 seconds
LLM analysis: 8-20 seconds (depends on signal count)
Formatting: <1 second
Total: 10-30 seconds per topic
Parallel Recap Generation:
Data Freshness
Signal Latency:
Total latency: 15-60 seconds from event to display (when refreshing actively)
Refresh Recommendations:
Breaking news: Every 2-5 minutes
Active monitoring: Every 15-30 minutes
Passive monitoring: Every 1-2 hours
Background awareness: 2-3 times per day
Rate Limits & Quotas
API Limits:
50 signals per feed per request
Unlimited refreshes (no rate limit on Y2 side)
Gloria AI may throttle excessive requests (>1 req/second)
Best Practice:
Don't refresh more than once per minute
Use AI Recap mode for historical signals
Live Feed mode for real-time only
Troubleshooting
No Signals Appearing
Problem: Terminal shows "No signals detected"
Solutions:
Check feed selection: At least one feed must be selected
Verify API connectivity: Gloria AI endpoint may be down
Expand timeframe: Some feeds have low activity
Try different feeds: Start with high-volume feeds (Bitcoin, AI)
Recaps Not Generating
Problem: "Generating AI recaps..." spins indefinitely
Solutions:
Check feed selection: No feeds selected = no recaps
Verify timeframe: Try 24h first (most reliable)
Reduce feed count: Generate 1-2 recaps first, then scale
Check quota: LLM API limits may be reached
Preferences Not Saving
Problem: Feed selections reset on page reload
Solutions:
Wait for save: Changes save after 1-2 second delay
Check authentication: Preferences require logged-in user
Clear cache: Browser cache may be stale
Check console: Database write errors may appear
Slow Performance
Problem: Terminal feels sluggish or unresponsive
Solutions:
Reduce selected feeds: 10+ feeds = more data to process
Use AI Recap mode: Fewer DOM elements than Live Feed
Limit Live Feed history: Only shows 50 most recent
Clear browser cache: Accumulated data may slow down
Mobile Experience
Responsive Design
The Signal Feed Terminal is fully optimized for mobile:
Feed Selection:
Horizontally scrollable feed pills
Larger touch targets (44px minimum)
Gradient indicators remain visible
Live Feed:
Single column layout
Condensed timestamps (e.g., "42m ago")
Full signal text (no truncation)
AI Recaps:
Full-width cards
Collapsed headers on small screens
Readable typography (16px base)
Mobile Workflow
Quick Check (2 minutes):
Open terminal on mobile
View pre-loaded data (from last session)
Scan recent signals
Refresh if needed
Deep Dive (10 minutes):
Switch to AI Recap mode
Select 24h timeframe
Generate recaps for key feeds
Read synthesized intelligence
Best Practices
For Active Traders
Setup:
Select 2-4 core feeds (e.g., Bitcoin, Ethereum, Macro)
Use Live Feed mode during trading hours
Keep terminal open in browser tab
Routine:
Refresh every 15 minutes during market hours
Watch for signal clusters (multiple signals on same event)
Cross-reference with price charts
Generate 12h recap at market close
For Researchers
Setup:
Select 6-10 feeds across domains
Primarily use AI Recap mode
Generate weekly recaps on Monday mornings
Routine:
Morning: 24h recaps for overnight catch-up
Midday: Live Feed spot check for breaking news
Evening: Note significant signals for weekly report
Weekly: 7d recaps for trend analysis
For Portfolio Managers
Setup:
Select feeds matching portfolio holdings
Use both modes (Live Feed + AI Recap)
Integrate with Y2 scheduled reports
Routine:
Pre-market: 24h recaps for all holdings
Intraday: Live Feed for breaking developments
Post-market: Generate 12h recap for daily summary
Weekly: 7d recaps for strategic review
Privacy & Data
What's Stored
User Preferences:
Selected feeds (array of strings)
View mode (news or recap)
Timeframe (12h, 24h, 3d, 7d)
Not Stored:
Signal content (ephemeral, from API)
Recap content (cached temporarily, not persisted)
User interaction logs (no analytics tracking)
Data Source
All signals originate from Gloria AI (itsgloria.ai):
Independent third-party intelligence provider
No Y2-specific customization
Same data available to all Gloria AI users
Y2 adds:
Multi-topic aggregation
AI-powered recap generation
Persistent preferences
Integration with Y2 intelligence reports
Last updated