← Blog
5 min read

Inside the Signal Radar: How We Built a Real-Time Intelligence Command Center

Top signals ranked by confidence, entity relationships mapped, story clusters with cascading effects, source bias visualization, live feed by vertical, and living briefs that get richer over time.

We rebuilt Signal Radar from the ground up. What was a simple feed of top signals is now a full intelligence command center — 7 live sections, all powered by existing API endpoints, auto-refreshing every 5 minutes.

One screen. Every signal that matters. Updated continuously.

What the Radar Shows

High Impact Signals

The 7 highest-confidence briefs across all 18 verticals, ranked by trust score. Each signal shows confidence bar, category, source count, and time since publication. Living briefs — ones that have been updated with new sources — get a glowing cyan indicator so you can spot evolving stories instantly.

Entity Graph

A visual node graph of the top 15 trending entities. Companies in cyan, people in green, countries in red, topics in yellow. Nodes are sized by mention count and connected by co-occurrence. Hover any node to see its connections light up and unrelated entities fade. On mobile, this becomes a ranked list.

Entity Momentum

A stock-ticker-style display of which entities are being mentioned most right now. Top 10 ranked by 24-hour mention volume with type badges and relative bars. Click any entity to see its full mention history and related briefs.

Signal Clusters

Related briefs grouped into story clusters with tabbed navigation. Each cluster shows its contributing signals with cascading effects — what happens downstream when these signals converge. This is how you spot emerging narratives before they become headlines.

Source Bias Map

Eight major outlets mapped on a left-right bias spectrum with proportional bars. Reuters and AP near center, CNN and NYT left-leaning, Fox News and WSJ right-leaning. At a glance, you see how the information landscape is framed.

Live Feed by Vertical

The 7 most recent briefs, filterable by category. Switch between Markets, AI, Crypto, Politics, and more with a single click. Auto-refreshes every 60 seconds. Briefs published in the last 5 minutes get a cyan “NEW” badge.

Category Pulse

All active verticals at a glance with health indicators. Green means high-confidence coverage, yellow means moderate, red means proceed with caution. Only categories with active briefs are shown — no empty rows cluttering the view.

Ask Polaris

At the bottom of the Radar is a terminal-style search bar. Press / from anywhere on the page to focus it. Type a question, hit Enter, and get matching briefs with a synthesis card — all without leaving the dashboard.

The placeholder rotates through example queries: “What signals affect AI infrastructure?”, “Track semiconductor supply chain risks”, “Show me energy market disruptions.”

Living Briefs

The most important thing about Signal Radar isn't any single section — it's that the briefs themselves are alive. When a new outlet covers the same story, Polaris doesn't create a duplicate. It updates the existing brief with the new source, recalculates confidence, and enriches the analysis.

You'll see this on any brief card with the ↻ LIVING BRIEF badge. A brief that started with 2 sources and 72% confidence might evolve to 12 sources and 94% confidence over a few hours. The brief detail page shows the full timeline: how many updates, when the last one happened, and the current source count.

Static articles are snapshots. Living briefs are the full picture.

Built for Agents Too

Everything on Signal Radar is available via API. Your agent can monitor the same signals programmatically:

python
from polaris_news import PolarisClient

client = PolarisClient(api_key="your-key")

# Get top signals by confidence
feed = client.feed(sort="confidence", per_page=7)
for brief in feed.briefs:
    living = f" [↻ {brief.update_count} updates]" if brief.update_count else ""
    print(f"[{brief.confidence:.0%}] {brief.headline}{living}")

# Get trending entities
trending = client.trending_entities(limit=15)
for entity in trending:
    print(f"{entity.name} ({entity.type}) — {entity.mentions_24h} mentions")

# Get only living briefs
living = client.feed(updates_only=True, sort="updated")
for brief in living.briefs:
    print(f"[{brief.source_count} sources] {brief.headline}")

Build alerts, dashboards, or autonomous monitoring agents on top of the same intelligence layer. The Radar is just one way to view it.

See It Live

Auto-refreshes every 5 minutes. Live feed updates every 60 seconds. No signup required.

Open Signal Radar →

What's Next