A Mastercard executive heard Laura Gebhardt, the company's VP of Global Product AI & Data Strategy, describe what AI agents could make possible and asked her to help his team rethink customer onboarding.
The team already had requirements and a solution in mind. They'd mapped the process and worked out a future state. From their perspective, the planning was well underway.
Laura shared the story at last week's AUTONOMOUS summit, hosted by Board of Innovation. Her session began with the work of opening that conversation back up.
Onboarding a customer involved multiple organizations and systems. Information was entered repeatedly. Each department ran its own review queue, with little visibility into the others, so the work moved sequentially by default. Employees chased updates and carried information between teams.
“People are the integration layer,” she said.
The leadership session had introduced a question broad enough to reconsider the entire experience:
“If intelligence and automation were everywhere, how would we redesign the experience from the ground up?”
Answering it required the people who understood how the work happened. They knew which cases needed judgment and how the departments depended on one another. Together, they developed a picture of the journey that went beyond the existing requirements. Laura described that work as part of a transformation still underway.
The model she outlined gives AI responsibility for coordinating work across the journey. People receive the context needed to make decisions, with clear authority over risk and exceptions.
BOI has now released the summit recordings, including Laura's Agentic AI at Scale: Lessons from Enterprise Transformation.
Watch Laura's session and explore the AUTONOMOUS recordings →
The business opportunity becomes easier to see at that scale. Onboarding affects when a customer can begin using a service and how much experienced attention it takes to get them there. Those factors belong in the economics of acquiring and serving customers.
They also reach across budgets. An improvement funded by one department can create much of its value somewhere else in the company.
The way an AI investment gets defined shapes what gets funded. A business case built around reducing the effort in one task can miss the value of changing the customer's whole experience. It can also leave out the work another team needs to do before that value becomes possible.
A company's AI portfolio can inherit the boundaries of its org chart.
An initiative enters through the function that proposes it. Its scope follows the work that team controls, and the return is calculated against that team's costs. Repeat that across the business and the investment plan starts to take shape around departmental improvements.
The larger growth opportunity may require several of those changes to work together.
Mastercard's onboarding example makes the dependency visible. Information collected at intake becomes more useful when reviewers can access it throughout the journey. A recommendation becomes actionable when the person responsible receives the relevant context and has clear authority to decide. Each improvement depends on what happens around it.
Take the path from customer insight to a product decision. A team could gain the ability to investigate an emerging need much sooner. Capturing that opportunity would also depend on how quickly the business can develop a credible concept and put it in front of customers. The investment has to account for the work needed to use the insight while it still matters.
My bet is on companies that make those connections part of their growth planning. They'll have a clearer basis for deciding where AI belongs because they've defined what the business needs to become capable of doing. They can evaluate the combined investment against a commercial outcome and put someone in charge of achieving it.
That changes the discussion about priorities. A shared data improvement could deserve funding because it enables several worthwhile initiatives. A promising pilot might need a change in another department before further development makes sense. An existing growth priority could require a larger redesign than the original project brief allowed.
Laura also addressed what happens when the organization's approval process moves more slowly than the work it is reviewing:
“Layers of approval quietly erase the speed advantage AI creates.”
Her point was that governance has to evolve alongside the transformation, with the business leading and technology working as a partner.
That gives the people funding AI another responsibility. They need to understand the relationships between initiatives and make decisions with the full business context available. Otherwise, each team has to reconstruct the larger case every time the work crosses a boundary.
What to do this week.
For the next portfolio review, put the growth plan beside the list of AI investments.
Choose one commercial priority and trace the initiatives contributing to it. Examine what the customer or business will be able to do once that work is live. Then identify the dependencies that could still prevent the outcome.
That exercise can reveal a missing investment or a project whose scope needs to change. It also establishes who is responsible for the combined result. Success needs to be measured where the value was promised.
From the portfolio
Within HauerX, AlignAI helps companies manage the decisions around their AI investments. It brings together each initiative's objectives and business context, supports evaluation against value, risk and feasibility, and routes the work through the relevant stakeholders. Leadership can follow initiatives through deployment and ongoing oversight.
The strategic work is deciding what the company is trying to accomplish. Keeping that purpose attached to the initiative as different teams evaluate and deliver it is part of making the investment work.
For a leadership team reviewing its AI portfolio, a useful starting point is one growth priority and the initiatives intended to support it. That's a conversation we can help bring together through HauerX.
What will your company be able to do when the AI initiatives you're funding work together?
