923
Occupations scored
18,796
Task-level analyses
1.2M+
AI benchmark evaluations
5.3M+
Worker survey annotations
What We Measure
Each occupation is broken into tasks and scored on three axes, using AI benchmark data and occupational surveys.
Structural Exposure
How rule-based and digitally accessible is the work? High scores mean AI can technically reach it.
Procedural Intensity + Digital Accessibility.
Human Imperative
Does the work require a human in the room? Physical presence, trust, empathy. Things AI can't fake.
Patients want human doctors. Clients want human advisors.
Demonstrated Capability
Can today's AI actually do the required skills? Based on benchmark results, not expert opinion.
1.2M+ evaluations across 95 models via Stanford HELM.
Every task lands in one of four zones: Automation, Augmentation, Human-Centric, or Status Quo.
What the API Does
Query jobs, tasks, skills, and regions through a REST API with a Python SDK.
Job-Level Scores
Automation susceptibility, human resilience, and net impact for every occupation.
Task-Level Breakdown
Break any job into its component tasks. Each scored on four dimensions with plain-language explanations.
Semantic Search
Find jobs, tasks, and skills by describing what you're looking for. No SOC codes needed.
Regional Data
Employment, wages, and exposure metrics across 31 California metros.
Career Transitions
Reskilling pathways with skill gap analysis and training priorities.
Custom Scoring
Weight the dimensions to match your priorities. Same data, your lens.
Who Uses Cubit
Enterprise HR, government policy, academic research, and financial analysis.
Enterprise Companies
AI exposure dashboards and reskilling plans across your full workforce.
Consulting Firms
Workforce strategy backed by data, not opinion.
HR Tech Platforms
Embed AI readiness scores into your workforce planning product.
Researchers
Structured, reproducible data for AI and labor economics research.
Government
Model employment impacts by region and industry.
Finance
Quantify workforce risk with real numbers, not headlines.
Built on Stanford HELM & O*NET · Validated against ILO, OpenAI & CAIS
Get Started
Free Sandbox access. No credit card required.