Financial/Macro · Score Gold
US Macro Economic Health Score
0 / 100
Stable
US · National
Confidence0.92
Live · monthly · next ~30d
Top drivers
- GDP Growth 0.40
- Unemployment Rate 0.35
⌁ mcp.call("adw-002") View →
Agent-ready intelligence objects
Stop rebuilding analysis.
Stop rebuilding analysis.
Start reasoning.
Alpine DataWorks delivers pre-computed intelligence objects — built on FRED, U.S. Census, BLS, NOAA, BEA and 20+ primary public sources — through one API and MCP endpoint your agents can reason over directly. The answer, with its drivers and confidence, already computed.
Incumbents sell raw data. Foundation models sell reasoning. We sell the explainable answer between them — the thing an agent actually calls. Stripe turned payments into one API call. Alpine does the same for intelligence.
Free tier — no card required. Discoverable by any MCP-enabled agent.
- 14
- build-ready products
- 13
- field IOM schema
- 100%
- free public data
agent · python
await adw.call("adw-002")
→ { score: 68, confidence: 0.92, drivers: [...] }
Built on primary public sources
FRED Federal Reserve
CENSUS U.S. Census Bureau
BLS Bureau of Labor Statistics
BEA Bureau of Economic Analysis
NOAA Natl. Oceanic & Atmospheric Admin.
CDC Centers for Disease Control
The honey hole
A $25–35B market sells data and dashboards.
A $25–35B market sells data and dashboards.
Almost nobody sells the answer.
2%
Of 2,773 public-data services are genuinely intelligent — pre-computed, explainable, agent-callable.
$25–35B
Serviceable market for the pre-computed answer. It exists. Nobody is selling it yet.
30 → 1
30 rows of raw data across 6 sources — collapsed into a single scored, driver-backed answer.
What travels with every answer
Not just a number. A position with its argument.
Drivers, not just data
Every object ships with its top drivers decomposed — you see exactly why the score moved, and by how much.
Confidence, not guesswork
A calibrated confidence level and freshness timestamp travel with every answer. No silent staleness.
Freshness, guaranteed
A clear refresh cadence and TTL per product. Stale data is flagged, never hidden, never silently served.
The refinery
From crude data to high-octane answer
Raw data is crude oil. Alpine is the precision refinery that outputs a standardized, agent-callable object — validated before it ships.
- 01
Source
Free public data — FRED, Census, NOAA, CDC, BLS, BEA — ingested and normalized with discipline.
- 02
Compute
Enrich, score, and decompose into drivers. Validated to beat a naive baseline before it ships.
- 03
Object
A single answer with its reasoning attached: score, drivers, confidence, freshness — one IOM schema.
- 04
Agent-call
One MCP/REST call returns the answer. No raw data to wrangle, no model to rebuild.
Intelligence products
Browse all products Flagship objects, ready to call
A taste of the 14 build-ready products. Each is a scored answer with drivers, confidence, and a freshness contract — callable today.
Financial/Macro · Score Gold
US Macro Economic Health Score
How healthy is the US macro economy right now?
Method: Z-score CPI/Unemployment/GDP; 1-(Inflation+Labor-Growth)
monthly · next ~30d score 68
View object → Market/Crypto · Index Gold
Crypto Market Sentiment & Volatility Index
What is the current sentiment & volatility regime for crypto?
Method: 0.4·Sentiment+0.3·Volume_Momentum+0.3·Volatility_Adjustment
hourly · next ~1h score 72
View object → Market/News · Index Gold
News Sentiment & Trend Velocity Index
Current global news sentiment & velocity?
Method: 0.6·Tone+0.4·Velocity
hourly · next ~1h score 55
View object → Retail/Location · Score Silver
Global Real Estate Affordability Score
How affordable is real estate vs income?
Method: 100-(Norm_Price_Income+Norm_Rent_Income)
monthly · next ~30d score 42
View object → Try it live
The object, the schema, the trend, the errors
Switch products, expand the IOM schema, see a trend, and simulate error states — the exact object your agents will call.
GET /v1/intelligence/adw-002
One object, four consumers
Built for humans, apps, marketplaces, and agents
The same typed object — same schema, same freshness contract — serves every consumer without re-integration work.
Humans
CFOs, CMOs, and operators get a decision-grade answer, not another dashboard to read.
Apps
Drop a scored index into your product with a typed schema and a freshness contract.
Marketplaces
List the same objects on data marketplaces and the MCP registry, unified schema.
Agents
Discoverable via /llms.txt + /catalog.json. The pre-computed answer an agent actually wants to call.
Receipts, not claims
We validate before we ship
No invented metric is sellable until it beats a naive baseline on real data. Our validated keepers carry their receipt — and our methodology is transparent without revealing the formulas.
Validated signal Validated
EWC — Earnings-Window Conviction
+0.125 incremental IC
Beats a naive baseline at p<0.05. One of the validated keepers cited on the methodology page.
FREDCensusNOAACDC
The refinery for intelligence — raw public data in, agent-ready answers out.
The answer your agents are waiting for is one call away.
Start with a free sample object. No card required for the Free tier.