THE MODEL
An organization designed around outcomes.
How AI agents change processes, structures, roles and capabilities, and how we design that change.
AI changes the unit of design.
Work
Analysis starts with which outcome to deliver, not how many people are needed. Tasks split into automatable, assisted, human and control.
Structure
Agents enter the process as execution capacity. Leadership coordinates outcomes, exceptions, risks and continuous improvement.
Talent
The relative value of transactions drops. Business judgment, integration, influence and learning rise.
Organization design, process redesign and talent strategy move as one system.
WHERE IS YOUR COMPANY?
Three questions, three levels of impact.
01 · Use
How do I do my task faster?
AI assistants and individual productivity. The role barely changes.
02 · Automate
Which part of the flow can AI do?
Agents that execute defined steps. Tasks and handoffs change.
03 · Redesign
What outcome can we deliver in a new way?
People and agents are designed together. Structure, decisions and talent change.
TOOL → WORKFLOW → OPERATING MODEL
The opportunity is not to automate today's organization. It is to redesign it around outcomes.
THE MODEL
From org chart to Outcome Cells.
In an agentic organization the unit of work is no longer the department: it is the cell. One business outcome, one owner, the people and agents who deliver it, and clear rules on who decides what.
Vertical flows · handoffs between areas · manual coordination
Cells by outcome
Cross-functional teams form around a business priority.
Talent in motion
Expertise moves across cells, not only up the reporting line.
Shared agents
One agent serves several cells and processes.
Explicit rules
Decisions, data, risk and accountability always visible.
How an Outcome Cell works.
01
Signals
Sales, POS, inventory, campaigns
02
Agent
Integrates data, detects anomalies, builds baseline
03
Person
Validates assumptions and scenarios
04
Agent
Proposes replenishment and simulates impact
05
Decision
Aligns service, cash, margin and risk
Outcome
One business KPI
Outcome owner
One accountable person
People
Who decide and resolve exceptions
Agents
Who run the routine
Rules
Decisions reserved for people
Data
The single source of truth
Human control: rules, data quality, approvals, risk and accountability.
The roles around each agent.
Outcome owner
Defines the outcome, business rules, priorities and value.
Capabilities
Business judgment, trade-offs, stakeholders
Process architect
Redesigns the end-to-end flow and removes handoffs.
Capabilities
Process excellence, controls and exceptions, continuous improvement
Agent owner
Manages the agent's performance, versions, limits and backlog.
Capabilities
AI literacy, orchestration, monitoring and quality
Human reviewer
Validates critical outputs and makes the decisions reserved for people.
Capabilities
Professional judgment, risk awareness, escalation
Change champion
Drives adoption, listens to friction and reinforces routines.
Capabilities
Facilitation, coaching, storytelling
Deliverable: role charter, RACI matrix, reserved human decisions and escalation criteria.
The 5 C's across five levels.
Draft · to be confirmed with HUF
| 01 · Explore | 02 · Apply | 03 · Integrate | 04 · Orchestrate | 05 · Multiply | |
|---|---|---|---|---|---|
| Commercial judgment | Understands how their work moves the P&L. | Uses business data to prioritize their work. | Connects area decisions to customer and market. | Defines outcomes and trade-offs for a cell. | Builds business judgment in other leaders. |
| Calculation | Reads basic indicators and dashboards. | Builds simple scenarios with agent support. | Decides with data, explicit assumptions and metrics. | Designs a cell's KPIs and simulations. | Installs a data-driven decision culture. |
| Connection | Uses AI assistants for own tasks. | Automates repetitive steps in their flow. | Redesigns end-to-end processes with agents. | Orchestrates several agents and their controls. | Scales reusable components across cells. |
| Convening | Shares learnings with their team. | Coordinates with other areas on exceptions. | Aligns stakeholders around an outcome. | Leads cross-functional mixed teams. | Mobilizes the organization toward the new model. |
| Change | Experiments with new tools. | Adopts new work routines. | Proposes improvements to the human-agent flow. | Manages adoption in their cell. | Develops champions and communities of practice. |
A career path toward the outcome.
- 01
Transactional executor
Work content
Captures, reconciles, prepares and follows up.
Success = Volume and accuracy
- 02
Augmented professional
Work content
Uses agents to speed up analysis and documentation.
Success = Quality and productivity
- 03
Business integrator
Work content
Connects data, process, customer and areas to resolve exceptions.
Success = Decisions and alignment
- 04
Outcome owner
Work content
Orchestrates people and agents; improves the system and owns the value.
Success = Impact and learning
Learn the tool → redesign the process → integrate decisions → lead outcomes
You don't launch an agent. You install a new way of working.
- 01
Case for change
An honest narrative: what problem we solve, what value we seek and what work changes.
- 02
Participatory design
Users and leaders co-design the human-agent flow, exceptions and controls.
- 03
Roles and accountability
Who configures, uses, supervises, approves, resolves exceptions and improves the agent.
- 04
Capabilities
Close gaps in AI, data, process, business, judgment and influence.
- 05
Adoption in the flow
Pilots, assisted practice, champions, team routines and visible feedback.
- 06
Governance and value
Guardrails, metrics, human decisions, risks and benefits.
Five agreements to start the redesign.
- 01OutcomeThe priority process and its KPI
- 02SponsorA leader accountable for value, risk and adoption
- 03TeamBusiness, process, IT, data, HR and control
- 04TalentCritical capabilities and the 70/20/10 learning model
- 05GovernanceHuman decisions, guardrails and success metrics
Let's design your first Outcome Cell.
Book a Sprint