LIVE · W16.26 · INDEXING 4,820 FIRMS · LAST REFRESH 17 MIN AGO

Unemployment data
is a lagging
indicator.

JobCreation.AI reads every open requisition at the world's largest employers — directly from public hiring channels — and turns hiring intent into a real-time signal. We see the hire before the headline, the layoff before the 8-K, the city shift before the quarter.

Labor flow is alpha.

Hiring decisions lead earnings by two-plus quarters. Layoff announcements are the end of a story the careers page has been telling for weeks. We turn that into a structured signal you can backtest, alert on, and overlay on your existing thesis.

  • 01Long/short baskets derived from net-hiring momentum across S&P 500 and sector sub-baskets.
  • 02Early-warning clusters — when three firms in the same SIC start cutting in the same week, you get pinged before the tape.
  • 03Lead-lag vs JOLTS / NFP with R² 0.87 over 104 weeks. Labor-flow prints three weeks before the bureau does.
  • 04AI-exposure long basket: firms whose req-level AI-tool mentions accelerated before their capex guide.
TRADE IDEAS · LABOR-FLOW DERIVEDLIVE · W16.26
SIDETICKERTHESISCONVICTION
LONGNVDAOpen reqs +18% w/w across CUDA + networking. Hiring precedes capex guide by ~2Q.◆◆◆◆◇
LONGORCLCloud infra reqs +8.4% w/w; backend + SRE adds mirror 2018 AZ-cycle.◆◆◆◇◇
SHORTINTCHillsboro eng cut confirms design-team attrition; F-500 churn at 24M high.◆◆◆◆◇
SHORTDISStreaming ops −420 and content reqs −18% YoY. Margin leverage turning.◆◆◆◇◇
LONGUBERPlatform eng +34% in Dublin + Amsterdam. Geo-arbitrage on seat cost.◆◆◇◇◇
§01THE SIGNAL STACK

Seven lenses on the same underlying labor flow.

Every open requisition in our dataset carries weekly history, semantic tags, and a stable color identity. Pivot the same data by tech stack, role, industry, geography, company, or AI adoption — bar, pie, or trend. Click anything to drill down. Share by URL.
LENS · 01

Tech Stack

213.4K ACTIVE · +4.8% W/W
LENS · 02

Roles

812 ROLES · 2.4M REQS
LENS · 03

Industry

24 SECTORS · 18 COUNTRIES
LENS · 04

Country

168 COUNTRIES · WEEKLY
LENS · 05

City

2,480 METROS · GEOCODED
LENS · 06

Company

F500 · DJIA · S&P · RUSSELL
LENS · 07

AI Adoption

2,840 MENTIONS · +31% W/W
+

Build your
own signal

SAVED · SHAREABLE
/signals/tech?country=IE
&seniority=staff+
&mode=trend
§02HOW IT GETS BUILT

From public signal to structured alpha, five steps.

We don't license third-party labor data. We build it from public sources, reconciled across channels, categorized, and versioned, so you can see not just what's posted today, but how it's drifted over 12 weeks.
STEP · 01
Ingest
4,820 firms · multiple sources each
STEP · 02
Reconcile
12.4% avg w/L mismatch
STEP · 03
Classify
14 axes · stable palette
STEP · 04
Version
W/W · YTD · YoY deltas
STEP · 05
Curate
ILO · WB · BLS anchored
§03EVIDENCE

Labor flow leads the BLS print by ~3 weeks.

LEAD / LAG · JC.AI LABOR-FLOW vs BLS NONFARM PAYROLLS · 12W

A real leading indicator, not a press release.

W05W08W11W14W17W20+3%+1%0−1%~3W LEADJC.AI LABOR-FLOWBLS NFP · LAGGED
R² 0.87
correlation
−3W
avg lead time
168
countries covered

You don't need more dashboards.
You need a leading indicator.

JOLTS prints monthly, six weeks after the fact. The 8-K lands the morning the story breaks. Earnings calls confirm what the market already priced.

A company's careers page is the single earliest public artifact of capital allocation decisions — before the announcement, before the filing, before the ticker moves. We index it, reconcile it, and make it queryable.

CadenceWeekly → real-time
SourcesPublic channels · multi-source
History12+ weeks per signal
APIREST · CSV · Atlas Chat
"The hiring page is the earliest public artifact of a company's next quarter. We just had to go read it."
§04AI POWERPLAY

The first real index of the AI industry.

DJIA and the S&P are exchange-listed. The AI stack isn't — yet. AI Powerplay is Intellibus's hand-curated universe of AI hardware, data centers, model providers, tool providers, and service providers, published publicly and powering a dedicated Signals lens.

AI Power Composite

Every AI company categorized across five layers of the stack, with self-serve listing corrections via two-stage curator-approved verification.

LAYER · 01
Hardware

GPU, TPU, custom silicon, networking fabric.

LAYER · 02
Data Centers

GPU-first operators running the compute fleet.

LAYER · 03
Model Providers

Foundation-model labs shipping weights + APIs.

LAYER · 04
Tool Providers

Products built on models — copilots, agents, search.

LAYER · 05
Service Providers

Consulting, integration, managed AI delivery.

INDEX
AI Power Industrial Average

Hand-picked basket — the DJIA of AI.

INDEX
AI Power 500

Top ~500 by composite methodology.

INDEX
AI Power 2000

Mid-cap breadth across the full stack.

INDEX
AI Power 3000

Deep-tail coverage — every serious AI name.