Editorial
White Paper
The Future of Leadership Development: AI-Driven Manager Enablement
In the previous article, we introduced an approach to redesign workflows for agentic AI. We now apply that approach to the travel spend workflow, a familiar enterprise use case.
The traveler’s experience exposes the limits of policy-driven processes when real-world conditions change. Travel spend is not broken because of policy gaps, but because edge cases such as delays, reroutes and geopolitical disruptions are now the norm. To understand where this breaks, we start with the current design.
Starting Point: Current Travel Spend Workflow
At Level 1, the current travel spend workflow is designed for a predictable travel journey in line with policy, with disruptions treated as exceptions after the fact. This typically results in a five-step flow with variable cycle times.
Reset the Starting Point: Simplify the Work
Instead of assuming the current flow is valid, we challenge each step in the workflow. The goal is to simplify or remove the work and redefine the outcome. When we remove the assumption of a stable travel plan, it enables a more responsive traveler experience under changing conditions, while applying policy through real-time, context-aware decisions.
The workflow structure then shifts from linear to adaptive. Adjustment is expected, and policy guides decisions in real time. The steps change, and cycle times are stabilized by minimizing exception handling and enabling decisions in the flow. Exceptions do not disappear but now represent true edge cases that require validation, rather than driving the workflow.
Reset the Starting Point: Simplify the Organization
At Level 2, redesign shifts focus from the work to the roles that own and make decisions about it. In the current state, the workflow involves five actors with fragmented ownership, three decision points and multiple handoffs that increase with disruption.
Now we challenge who does the work and where decisions are made, with the goal of consolidating ownership and accountability while supporting distributed execution. Decisions move closer to where the work is done.
The redesigned model reduces the number of actors to a single accountable owner at the point of travel. Decision points shift into the new flow of work, and handoffs are minimized. Exceptions are not routed across teams but addressed within the flow, with edge cases validated at settlement.
Financial oversight shifts from active control to post-travel validation. Trust is designed into the workflow through real-time policy guidance, removing the need for escalation loops.
Reset the Starting Point: Simplify the Architecture
At Level 3, the focus is on the systems and data that enable the work. Capturing this context is essential to executing the workflow under changing conditions.
In the current state, the travel spend workflow relies on five core systems and their data sources. These include travel booking, expense management, workforce management (HR), finance and productivity systems, with data created, read and updated across each step.
Any disruption triggers changes that must be synchronized across these systems, which introduces complexity and the need to reconcile and align data.
By reducing the number of systems and data dependencies, we simplify the architecture and eliminate the need for coordination in the flow. This is achieved by introducing an orchestration layer and a shared data resource, creating a unified operational context within which the end-to-end workflow can execute with greater consistency and speed.
This approach simplifies the architecture from five systems to two core platforms, supported by a unified data source and a concierge-led experience. The role of systems shifts from coordinating the workflow to supporting it within the shared context, where triggers and updates are handled in real-time.
Decide Before You Automate
By automating the current state, agents flag deviations, route exceptions and coordinate across systems after the fact. This approach increases complexity and adds to the human cognitive load with validation of agent output.
The redesigned workflow changes the sequence. Work is simplified, ownership is consolidated, and architecture is unified before automation is introduced. Automation is then applied selectively based on the nature of the work and defined context.
- Work that has been eliminated is not automated
- Decisions that require judgment remain guided, not delegated
- Execution steps become candidates for automation only after context, ownership, and dependencies are clearly defined
Deploying agents before clearly defining Levels 1–3 creates workflow debt and limits ROI as automation pre-requisites of outcomes, ownership and context are not met. This example uses an ultralightweight model for illustration, without the need for complex tooling.
Communicate and Calibrate Through Learning
Implementation is only the start. The redesign only becomes real through active, simple communication grounded in the work. The goal is to help people understand what is changing and why, and to reduce attachment to the current way of working. Adoption does not require agreement. It requires clarity and practice.
Communication is anchored in the Level 3 process map and focused on what has changed. Start with the business problem: variable cycle times, finance workloads and friction, and a fragmented travel experience. The redesign simplifies the work, organization and architecture to improve all three. Progress is not defined by rollout, but by how quickly the workflow is understood, used, and improved in practice.
This means:
- Refer to the map when explaining changes: what moved, what simplified, what stopped
- Review key metrics at a defined cadence and adjust the workflow based on usage
- Avoid creating new artifacts or documentation layers
- Anchor communication in what is observable
Before: five steps with variable cycle times, five actors with three decision points and multiple ownership handoffs, five core systems and data dependencies that require synchronization.
After: four steps, a single accountable owner, and two core platforms supporting execution.
This exercise tests the redesign model against real work. It quickly became clear that this is not just a design exercise. Stay anchored in the business problem, resist the urge to over-engineer, and capture context as it exists in practice.
Engaged, But Not Convinced: What's Eroding Beneath Your Strongest Engagement Numbers
Watch Now
Retention Is a Communication Problem: New Data from 6,000+ Employees
Watch Now
Do Learning Programs Really Work? How to Turn Education Into Engagement In Healthcare
Watch Now
Fuel Your Front Line: Workforce Trends Shaping 2026
Watch Now
The model provides a starting point, but it is not complete on its own. As workflows evolve in use, context decisions and patterns will continue to shift and require refinement. This includes how agents behave at runtime, how response accuracy is assessed, and how learning is incorporated back into the workflow.
Please continue to share your use cases and feedback so we can test and learn together.
Editor's Note: Catch up on Ro's previous series on digital fluency below:
- Why Digital Fluency Matters More Than Basic Digital Literacy — Digital fluency gives employees the mindset and skills to lead, not just follow — capabilities once reserved for executives but now essential from day one.
- Digital Fluency: Turning Business Acumen Into a Strategic Advantage — Business acumen is the missing link between digital fluency and AI success — here’s how leaders can close the strategy gap in 2026.
- Digital Fluency: Designing Intelligent, Human-Centered Workflows — Workflow redesign is the single strongest predictor of scaled value from AI. When done intentionally, it helps employees and serves the business's bottom line.
Learn how you can join our contributor community.