01 · The framework
Integrated Organizational Capability and Change Framework
A holistic view of organizational design across four interconnected dimensions around a leadership core. Start with the overview, then select any quadrant to explore its components, playbook modules, and diagnostic questions.
Accountabilities
Skills
Behaviors
Systems
The overarching idea
One integrated system, not four separate fixes
Most transformations stall because they treat organizational design as a sequence of isolated moves — a new structure here, a training program there, a values poster on the wall. This framework treats capability as a single integrated system: four interdependent domains, held together by leadership and expressed through the organization's culture values.
Leadership sits at the center as the integrating force — aligning the four domains rather than optimizing any one in isolation. Culture values form the outer ring: the shared beliefs every domain operates within.
Change lands only when all four move together — which is why the change model activates every domain at once instead of running them as phases. Select a quadrant to explore its components, playbook modules, and the diagnostic questions that measure it — or take the diagnostic to see how your organization scores across the whole system.
02 · The five activators
What turns a design into performance
These five activators bridge organization design and results — the right connections, the right conversations, and the right know-how.
03 · The change model
The three C's of change
An operating system for driving organizational change — moving from awareness to embedding new behaviors. Each lever activates all four framework dimensions at once; they are not sequential phases.
1 · Narrative
Communicate
Create compelling narratives and visual representations of the desired future state. Use storytelling and imagery to help people see themselves in the new organization.
GoalMove from 'telling' to 'visualizing'
MethodDestination Postcard - visual blueprints that paint the future state
ToolTransformation Portal (single source of truth)
2 · Engagement
Co-Create
Engage employees at all levels in designing the new organization. Use rapid prototyping and iteration to test ideas. Align with Agile Change Management principles.
GoalEnlist stakeholders to reduce resistance
MethodScenario-Based Sprints - stress-test designs before go-live
3 · Reinforcement
Cadence
Establish regular rhythms and rituals that reinforce desired behaviors. Use quick feedback loops to maintain momentum and course-correct as needed.
GoalAlign goals and embed new behaviors into daily rhythm
MethodGoal-setting alignment; T.C.C.A.R. Integration (Trust, Conflict, Commitment, Accountability, Results)
ToolFlash Pulse Checks after key meetings
Cadence tool
The T.C.C.A.R. team health lens
Trust, Conflict, Commitment, Accountability, Results — based on Patrick Lencioni's five dysfunctions model, adapted for organizational design. Sustainable cadence depends on which of these five foundations your team can rely on.
Vulnerability-based trust where team members feel safe to be honest and authentic
Healthy behaviors
- ✓Open sharing of mistakes and learning
- ✓Asking for help without fear
- ✓Giving benefit of the doubt
- ✓Constructive feedback is welcomed
Warning signs
- ×Hiding weaknesses from one another
- ×Hesitating to ask for or offer help
- ×Holding grudges and suspicion
- ×Dreading meetings and avoiding interaction
Quick check: which dimension does your team struggle with most? Start there — trust is the foundation the other four are built on.
04 · AI-native design
AI redesigns the work, not the chart
Structure follows work — and work is now shared with machines. Three moves make an AI-era design defensible: descend to the task, place every task on the ladder, and let no machine judgement reach a decision-maker unreviewed.
Move one
Descend to the task
Organisation
Layers, spans, anchor layer — where accountability sits
Position
Grade, cost, reporting line — what HR systems can move
Task
Where AI attaches — scored and reassigned between human and machine
Process
The value stream those tasks assemble into, across every box
Findings roll back up the org graph and across value streams — that is how a task-level insight becomes a layer, function and P&L conversation.
Move three
Gate the machine's judgement
Every score and mapping carries a source and a confidence value. Low confidence is routed to human review by score, not by source label — and approvals become the house standard.
Human-in-the-loop is a workflow with a gate, not a disclaimer at the back of the deck.
Move two
Place every task on the ladder
Human end to end. Judgement, presence or dexterity is the point.
Nothing. Protect it — and stop counting it as a laggard.
Same task, faster — drafting, summarising, retrieval, first-pass analysis.
Tools and habits. No structural change; quiet productivity gains.
The machine runs the task; the human sets intent and handles exceptions.
Roles are rewritten and spans widen. Supervision replaces execution.
The task stops existing in its current form; the process is redrawn around the outcome.
An operating-model decision — layers, hand-offs and decision rights move.
Capability is only half the question
What AI can do, and whether the people doing the work want it automated — scored separately, never collapsed.
Green zone
Wanted and feasible. Start here.
Red zone
Feasible but resisted. A change problem, not a build problem.
R&D frontier
Wanted, not yet possible. Watch and pilot.
Low priority
Neither wanted nor feasible. Leave it alone.
Physical work carries no software-agent overlay — it surfaces as a separate robotics and equipment lever, so the analysis never quietly automates a ramp.
Provenance — every claim carries its source
Regulator-backed
National skills frameworks. Quotable to a regulator or a union.
Research-backed
Published task and agency datasets. Foreign preference data stays a hypothesis until tested locally.
Model-inferred
Useful, fast — and never decision-facing until a human signs it.
05 · The diagnostic
Ready to assess your organization?
Eighteen questions across five dimensions, in about five minutes — a baseline read of organizational health, where to focus first, and a targeted intervention for each weak spot.
About the author
Eric Yim
Organization Design Strategist
Global Head of OD and Talent, SATS Ltd. · Singapore
Eric Yim is an organization design strategist specializing in human-AI work partnership — designing how organizations divide work between people and machines, and the structures, capabilities and behaviors that make that partnership work.
25 years in organizational development — across Shell, Cargill, L’Oréal and the Singapore public sector.
Nanyang Technological University · 1998 · University of North Texas · 2022 · Certified Prompt Engineer™
- Organization Design
- Human-AI Work Partnership
- Operating Model Design
- Change Management
- Job Redesign
- AI Adoption in the Workplace
Profiles & coverage
“My life mission is to build effective organisations.”
Common questions
Who is Eric Yim?
Eric Yim is an organization design strategist specializing in human-AI work partnership. He has spent 25 years working between organizational diagnosis and redesign, and works at the intersection of organizational development and AI adoption in the workplace.
What is human-AI work partnership?
Human-AI work partnership is the deliberate design of how people and machines share work at the task level. Rather than asking which jobs AI replaces, it asks which tasks move to the machine, which stay human, and what that reassignment does to roles, spans, decision rights and the operating model.
What does an organization design strategist do?
An organization design strategist aligns an organization’s structure and accountabilities, people and skills, process and systems, and mindset and behaviors so it can execute its strategy. The work spans diagnosis, operating model design, and the change activation needed to make a new design real.
How does AI change organization design?
AI redesigns the work, not the chart. The unit of analysis descends from the org chart to the task, each task is placed on a ladder from manual to reinvention, and any machine judgement is gated by a human before it reaches a decision-maker. Structure follows work — and the work is now shared with machines.
What is Eric Yim’s background?
Eric Yim graduated from Nanyang Technological University in 1998 with a Bachelor of Business specializing in financial analysis, and began his career in corporate banking at Mizuho. He moved into organizational development through the Civil Service College, then held organization development and effectiveness roles at L’Oréal, Shell and Cargill before becoming Global Head of OD and Talent at SATS Ltd.
Where does Eric Yim work?
Eric Yim is Global Head of OD and Talent at SATS Ltd. in Singapore, which he joined in December 2024. He previously spent over nine years at Shell in global and regional organizational development leadership roles, and four years at Cargill in organization effectiveness consulting.
Why does Eric Yim say AI is a multiplier and not a shortcut?
Because a multiplier amplifies whatever expertise already exists — and zero multiplied by any number is still zero. AI is most valuable to people with deep domain expertise to amplify, not to those merely quickest to adopt the tools. The discipline is discerning the right time and place to use AI.