AI Workforce Diagnostic Platform

Every AI Forecast Your Board Has Seen Is Wrong

The Scaffold diagnoses how much of your team’s work is automatable, maps exposure by department, and builds 90-day transformation plans — grounded in O-Ring complementarity theory, not replacement math.

The O-Ring Error in Workforce Planning

McKinsey projects 30% productivity gains. Goldman Sachs forecasts widespread role elimination. These projections share a foundational flaw: they model tasks as substitutes. If AI handles 60% of a role’s tasks, the role becomes 60% more productive.

But Gans and Goldfarb’s O-Ring theory, confirmed by Anthropic’s Economic Index, shows tasks are complements that multiply. When you apply the correct math, projected gains drop by 50–67%.

This isn’t a minor adjustment. It’s the difference between a workforce strategy that works and one that fails.

The Problem

Excellent Macro Research. Zero Micro Methodology.

What Research Tells You

  • OpenAI’s GDPval maps 19,265 tasks across 1,016 occupations
  • Anthropic’s Economic Index confirms task complementarity
  • Microsoft’s AI Diffusion Index tracks real-world adoption at 16.3% globally
  • Gans-Goldfarb provides the O-Ring theoretical framework

What Research Can’t Tell You

  • Which of YOUR roles have valuable judgment bottlenecks vs commodity residual
  • Whether AI will superskill your teams or deskill them
  • Which departments are transformation priorities and which are already resilient
  • Where your specific competitive advantage sits on the pattern-judgment spectrum

This diagnostic gap — between macro findings and micro workforce decisions — is where organizations are flying blind. The Scaffold fills it.

The Mirror Moment

Your People Believe 40% of Their Work Is Pattern-Based. The Reality Is 60–80%.

This 20–40 point perception gap is what we call the Mirror Moment. Every knowledge worker will face it. The only question is whether they face it with organizational support — or market brutality.

“The most damaging thing you can do is let the market reveal their work composition to them. The most empathetic thing you can do is help them understand it first.”
20–40pt
Perception Gap
60–80%
Actual Pattern Work
50–67%
Projection Overestimate
4
Converging Research Streams
The Methodology

Four Dimensions. Five Research Streams. One Diagnostic.

The PJRC framework isn’t invented theory. It’s the synthesis of five converging research efforts that independently arrived at the same conclusion.

Pattern

Systematizable work where known frameworks apply and inputs predictably map to outputs. The domain AI handles first and best.

Judgment

Contextual decisions requiring interpretation of novel situations with multiple valid outcomes. The bottleneck that concentrates your value.

Relationship

Trust-dependent work centered on stakeholder alignment, negotiation, and human connection that cannot be transferred through documentation.

Creativity

Genuinely novel synthesis and strategic insight. Not template-filling disguised as creation, but true cross-domain connection.

The ratio between these dimensions determines whether AI superskills your workforce or deskills it. Until you measure it, you’re guessing.

Organizational Intelligence

From Individual Diagnostic to Organizational Strategy

Each assessment is a data point. Aggregated across your organization, they become a strategic map of transformation priority and competitive advantage.

01

Assess

Conversational AI assessment powered by Claude. Four-phase interview — Context, Time, Cognitive, Identity — produces a PJRC breakdown for each role. No forms. Real dialogue adapted in real time.

02

Map

Aggregate assessments into organizational heat maps. See automation exposure by department, skill band distributions, and pattern work percentages. Identify which teams face superskilling trajectories and which face deskilling risk.

03

Act

Department-level transformation roadmaps prioritized by exposure. 90-day plans with weekly check-ins. Before/after dashboards tracking the shift from pattern work to judgment work across your organization.

Heat maps of automation exposure by department and skill band
Superskilling vs deskilling trajectory identification by team
K-anonymity privacy protection for all aggregated data
Integration with O*NET’s 1,016 occupations and 18,797+ task statements
For Individuals

AI Doesn’t Threaten Your Value. It Concentrates It.

For workers whose roles contain judgment bottlenecks, AI removes the pattern work diluting their impact. The bank teller was always a relationship manager forced to count cash.

You might discover…

  • The 30% of your role that creates 80% of your value
  • Which skills to double down on and which to let AI handle
  • Where you sit on the Executor-to-Architect progression
  • A 90-day roadmap from recognition to reinvention

How it works for you

  • Natural conversation with our AI Assessment Conductor — no forms
  • Personalized PJRC breakdown backed by O*NET labor data
  • Skill band classification with research-backed comparisons
  • Transformation plan with weekly check-ins and progress tracking
Start Your Individual Assessment

Whether you’re leading an organization or navigating your own career, every insight starts with a single conversation.

The Evidence

Built on the Convergence of Five Research Streams

Gans-Goldfarb O-Ring Theory

Tasks as complements, not substitutes. The mathematical foundation showing standard AI projections systematically overestimate.

OpenAI GDPval

19,265 task-level assessments across 1,016 occupations. The most comprehensive mapping of AI capability to human work.

Anthropic Economic Index

Real-world confirmation of task complementarity. The first large-scale data showing the 50% discount when modeling tasks correctly.

Microsoft AI Diffusion Index

Enterprise adoption data showing where AI is actually being deployed, not where analysts project it will be.

O*NET Labor Framework

U.S. Department of Labor occupational database. 1,016 occupations, 18,797+ task statements — the backbone of workforce classification.

The Scaffold doesn’t compete with these research efforts. It operationalizes them.

1,016
Occupations Analyzed
18,797+
Task Statements Mapped
500+
Professionals Assessed
5
Research Streams Synthesized

The Pattern Paradox

“The organizations most vulnerable to AI disruption are the ones that succeeded by scaling through standardization. Their competitive advantage — efficient, repeatable processes — created exactly the pattern work AI handles best.”
From “The Work That Remains” — the research behind The Scaffold

The Market Won’t Wait. Neither Should You.

Every week without workforce composition data is a week your AI strategy is built on assumptions.

Individual assessments take ~25 minutes. Organizational pilots typically start with 10–15 assessments.