The Framework
The intellectual foundation for managing workforces that are no longer entirely human.
The Gap Is Not Technology. It's Framework.
The world of work is transforming. Knowledge workers — in every function, in every industry — now work alongside AI agents that are not just tools but participants in the work itself. Agents that carry capacity, handle volume, produce outputs, make decisions within defined boundaries, and require the same ongoing management attention that human workers require.
Workforce intelligence — with its focus on supply and demand, capacity optimization, and the operational systems that connect them — is the natural intellectual foundation for navigating this transition. Not because one discipline should own it, but because the frameworks built under operational pressure are exactly what every function managing workers — human and AI — now needs.
“You cannot reach Level 4 by doing Level 3 better. The organizations that make the jump do so by redesigning their workforce planning function around the blended workforce.”
— WFM Labs Thesis, 2026
The Three-Pool Model
Every organization deploying AI alongside humans faces the same structural question: how does work flow between human and machine capability? The three-pool model provides the answer.
Autonomous AI
AI agents handling work end-to-end. Elastic scale — expands in minutes, not weeks. Linear cost structure at a fraction of human productive hour cost. The constraint is capability, not throughput.
Staffing Model: Cost per transaction + escalation cascade costs + maintenance decay
Management: Budget, architecture, knowledge systems, model updates
Key Risk: Escalation tax — a $0.20 AI interaction with 30% escalation has an expected total cost of $2.06
Collaborative Portfolio
The most operationally novel contribution. Humans oversee concurrent AI sessions — 5 to 7 simultaneously — intervening at judgment boundaries. This is air traffic control, not traditional management. A controller doesn't fly the planes. They manage a portfolio of concurrent situations, intervening when it matters.
Staffing Model: N-star cognitive load model — capacity determined by concurrent oversight ability, not throughput
Management: Cognitive capacity monitoring, judgment quality, intervention timing
Key Insight: This role doesn't exist in most organizations today. It represents 5–7x leverage.
Human Specialist
Complex, high-stakes, relationship-driven work requiring full human expertise. Not disappearing — transforming. As AI absorbs routine work, what remains is harder, more consequential, and more emotionally demanding.
Staffing Model: Traditional Erlang-C + complexity premium (1.18x at 50% automation)
Management: Deep expertise development, emotional resilience, decision authority
Key Insight: A single empathetic specialist interaction generated $67,000 in customer lifetime value. This is the work worth investing in.
Five Levels of Workforce Intelligence
Most organizations operate at Level 1 or 2. The transformation to Level 4 requires a fundamentally different conception of what workforce management is for.
Manual
Spreadsheet-driven, reactive. Workforce = headcount.
Unlocks: React faster, fewer scheduling errors.
Foundational
Deterministic planning. Erlang-C. Pattern-based forecasting.
Unlocks: Predictable service, cost control.
Progressive
Continuous planning. Intraday automation. Variance as signal, not enemy.
Unlocks: Adaptive capacity, real-time responsiveness.
Phase Transition
Levels 1–3 are progressive improvement within the same paradigm. Level 4 is a structural redesign.
Advanced
The ecosystem emerges. Specialized OR planning engines exchange bidirectional data with core WFM. Monte Carlo simulation, multi-objective optimization, and evergreen planning replace seasonal cycles.
Unlocks: Maximum value per workforce dollar through multi-objective optimization.
Pioneering
Enterprise-wide intelligence. Complete integration of WFM into enterprise decision-making. Adaptive ecosystem optimizing human potential. Predictive operations sense demand before it arrives.
Unlocks: Competitive advantage through workforce intelligence.
The Second-Order Effects
Traditional planning ignores four effects that determine whether AI deployment actually saves money. 40% containment does not equal 40% savings. Expect 20–25%.
Escalation Tax
Remaining work is harder. Cycle times increase even as volume drops.
Rebound Demand (Jevons Paradox)
Total demand grows as service improves. Better AI = more people requesting things they wouldn’t have bothered with before.
Complexity Concentration
Hard work clusters. As easy interactions are absorbed by AI, what remains requires deeper expertise and longer cycle times.
Savings Erosion
Containment rates decay over 12 months. Without active lifecycle management, AI agents degrade — they’re frozen in time while the business changes.
The Future Belongs to Organizations That Design This Deliberately
We’re building the frameworks, the tools, and the community to navigate workforce transformation. Join us.