Published by Roast & Rise
The Work Chart: Designing Your AI-Native Org
Redraw one department around outcomes, agent work, human judgment, and clear ownership.
Build a one-page work chart that shows what your department must achieve, where agents execute, and where humans direct or intervene. Then test the structure with your team and improve it using two weeks of real work.
Course thesis
Agents change the structure of work because execution volume can move into agent lanes while humans retain direction, judgment, and outcome ownership. A work chart makes that shift visible through outcomes, lanes, owners, handoffs, and oversight roles. The Microsoft, CIO, Gartner, and McKinsey claims supplied in the brief are directional grounding; exact language and figures must be verified against the original sources before publication.
What you leave with
By the end, you will have a reviewed one-page work chart, a named human owner for every agent, and clear role cards for the four oversight responsibilities.
For
Founders, executives, and operations leads responsible for redesigning a department where AI agents already perform meaningful execution work.
Workflow
Redesign one department by mapping its outcomes, separating human and agent work lanes, naming a human owner for every agent, assigning oversight responsibilities, reviewing the structure with the team, and revising it after two weeks of live use.
Change
The reader replaces a headcount-led org chart with a one-page work chart for one department, reviews it with the team, runs it for two weeks, and revises it using real ownership and exception evidence.
What you can do
Use these as checks while you move through the plan.
Map one department around owned outcomes rather than reporting boxes.
Separate agent execution from work requiring human judgment, direction, or consequential review.
Give every deployed agent a named human owner and an explicit escalation route.
Assign the four oversight responsibilities and refine the structure using two weeks of operating evidence.
Chapters
01
Map the Work Behind the Boxes
Map one department through owned outcomes and the recurring work behind them, then expose gaps before redesigning its structure.
An org chart reveals reporting lines. It hides the work moving through them. Before assigning agents or changing roles, expose the department’s real work surface: the outcomes it owns and the recurring work required to produce them.
Start with five to eight outcomes. Write each as an observable result the department can influence and accept accountability for. Revenue, customer response time, accurate reporting, or resolved incidents can qualify when the department genuinely owns them. Broad activities such as “support sales” or “manage operations” need sharper boundaries.
Connect recurring work to each outcome. Use work areas rather than individual tasks, so the map stays at department level. Record volume, judgment demand, and exception risk. These signals prepare the next chapter’s lane decisions without making them early. Estimates are acceptable when data is unavailable, provided they are labelled as assumptions to validate.
The gaps matter. Work with no outcome may be legacy activity, hidden infrastructure, or waste. An outcome with no supporting work may expose a missing capability. Shared or unnamed ownership signals a structural decision waiting to happen.
Quality checklist
Outcomes describe observable results.
Every recurring work area has a visible connection.
Ownership gaps are clearly marked.
Unknowns are labelled for validation.
Common mistakes
Using job descriptions as the source of truth.
Breaking work into task-level detail.
Treating estimated volume as measured fact.
Hiding duplicated work under broad labels.
Checkpoint
Can every recurring work area be traced to an owned outcome, with every uncertainty visibly marked?
Exercise
Expose the Department’s Work Surface
- Choose one department and define what sits inside its boundary.
- Write five to eight observable outcomes the department owns.
- Group recurring work beneath each outcome.
- Mark volume, judgment demand, exception risk, and current ownership.
- Flag orphaned work, unsupported outcomes, and assumptions for team validation.
Use this at work tomorrow
Ask the department lead to mark one work area whose value or ownership remains unclear.
02
Draw the Work Chart
Allocate department work into explicit human, agent, and shared lanes while keeping every outcome and agent under named human accountability.
Your outcome map shows what the department must deliver. The work chart now makes execution and accountability visible.
Allocate each work area using four signals: repeatability, judgment demand, consequence, and exception frequency. Stable, repeatable work may enter an agent lane when capability has been validated. Work that sets direction, resolves ambiguity, or carries serious consequences stays in a human lane. Where execution crosses between them, define a shared handoff with a trigger, receiver, and required response.
These lanes describe operating boundaries. They are not permanent verdicts on what an agent can do. Record uncertain capability or risk as an assumption to validate during live use.
Keep accountability human and singular. Every department outcome needs one accountable human. Every deployed agent also needs one named human owner who understands its scope, monitors performance, and receives escalations. The agent may execute substantial volume. Its owner remains answerable for how that execution serves the outcome.
Quality checklist
Every lane connects to a mapped outcome.
Each outcome has one accountable human.
Every deployed agent has a named owner.
Handoffs specify trigger, receiver, and response.
Common mistakes
Allocating work from job titles instead of the outcome map.
Treating current agent capability as permanent.
Using “the team” as the accountable owner.
Leaving shared work without a handoff trigger.
Checkpoint
Can someone trace every outcome through its lanes and handoffs to one accountable human?
Exercise
Draw One Department’s Work Chart
- Copy each outcome and recurring work area from your department outcome map.
- Score each work area for repeatability, judgment demand, consequence, and exception frequency.
- Place it in a human lane, agent lane, or shared handoff. Mark uncertain placements as assumptions to validate.
- Name one accountable human per outcome and one human owner per deployed agent.
- Compress the result into a one-page work chart with an ownership register.
Use this at work tomorrow
Ask each named agent owner to confirm the agent’s operating boundary and escalation trigger.
03
Run the Oversight Loop
Assign the four oversight responsibilities, test the work chart in live use, and revise its structure from logged evidence.
Your work chart becomes credible when it survives contact with real work. The four oversight responsibilities turn the chart from a design into an operating system.
The agent supervisor watches day-to-day performance and boundaries. The eval owner defines how quality is tested and decides when evidence is strong enough. The exception handler takes work that falls outside the expected path. The human-in-the-loop reviewer approves, rejects, or revises consequential outputs before they move forward.
These are responsibilities, not mandatory job titles. One person may hold more than one role when workload and risk allow. Each role still needs a named assignee, decision rights, escalation triggers, and a destination for escalation. Otherwise the team sees responsibility without knowing who can act.
Run the chart for two weeks. Capture structural evidence: exceptions, failed handoffs, review load, and moments when ownership becomes unclear. Review the log with the people doing the work. Change the chart where the evidence reveals a weak lane, missing boundary, overloaded review point, or ambiguous owner.
Quality checklist
All four responsibilities have named assignees.
Each role card grants a concrete decision right.
Every escalation trigger names a destination.
Chart revisions cite dated operating evidence.
Common mistakes
Turning responsibilities into titles without authority.
Letting exceptions disappear into private messages.
Changing the chart after one unusual incident.
Assuming current agent capability will remain stable.
Checkpoint
Can every team member name who acts when an agent fails, drifts, or produces a consequential output?
Exercise
Put the Work Chart Under Pressure
- Assign one person to each oversight responsibility on your existing work chart.
- Give every assignee a decision right, escalation trigger, and escalation destination.
- Open a shared operating log and schedule the team review for 14 days from launch.
- During the trial, record exceptions, failed handoffs, review load, and ownership gaps.
- At the review, revise the chart only where logged evidence supports a structural change.
Use this at work tomorrow
Open the shared operating log and ask the four oversight assignees to confirm their decision rights.
30-day path
Days 1 to 3: Select one department, define its boundary, and name the design owner.
Days 4 to 7: Map department outcomes and connect current work to them. Flag gaps, duplication, and orphaned work.
Days 8 to 10: Allocate human lanes, agent lanes, and shared handoffs. Record assumptions that need live validation.
Days 11 to 14: Name a human owner for every agent, adapt the four oversight role cards, and review the full chart with the team.
Days 15 to 28: Run the structure. Log exceptions, unclear decisions, failed handoffs, review load, and owner confusion.
Days 29 to 30: Revise the work chart and ownership register. Confirm decisions with the team and set the next review date.
Success signals
By day 14, one department has a complete one-page work chart reviewed with the people doing the work.
Every deployed agent in scope has one named human owner in the ownership register.
All four oversight responsibilities have named assignees, decision rights, and escalation triggers.
After two weeks of live use, the chart is revised using logged exceptions, failed handoffs, review load, and ownership gaps.
Reflection prompts
Where does this topic show up in real work?
What behavior should change first?
What evidence would prove this Riseplan worked?
Manager checklist
Choose one owner for the behavior change.
Use the exercise on live work.
Review the output before scaling the habit.
Decide what changes after 30 days.
In this library
Related RisePlans
Agentic Work Redesign Sprint
Learn how to redesign work for teams using AI agents to handle delegated, long-running, and cross-functional tasks. Build a new operating model that makes room for parallel delegation, reusable instructions, modern review cycles, and the next level of team collaboration.
From Chatbots to Superagents
Stop settling for one-off chatbot interactions. This plan shows you how to delegate real, repeatable work to AI agents with control, confidence, and results your team can trust.
AI Search Visibility: A Practical Sprint
Audit how your company appears in AI search, publish verifiable source pages, set crawler policy, and measure changes in Search Console.
Want this shaped around your company?
Risey can research your company foundation first, then build a version of this path around your real workflows, customers, and culture.
Start with your company