Getting AI-ready: How Ochsner Health stood up EA and identified $16M in savings with Ardoq

Board-Level Imperative to Drive AI Innovation Amongst Rising IT Costs, Tech Debt, and Aging Infrastructure

Aaron Buck didn't arrive as CTO at Ochsner Health from a career in healthcare. After a pre-med start derailed by his own brain aneurysm at 17 and four years in a Fort Worth emergency room, he moved into technology, spending two decades in IT at large industrial organizations. The chance to return to medicine in a role that could help save lives was what drew him to Ochsner Health.

What he walked into was a gap between ambition and reality. The board had high AI aspirations and wanted to change healthcare across Louisiana and the Gulf Coast. When Aaron looked at whether the organization could deliver on that, he recognized that modernization would need to be the first step towards closing the gap.

Aaron's assessment surfaced the things that needed attention first: aging network infrastructure approaching end of support at a third of locations, and data centers that weren't yet meeting the tier-one standard the organization was aiming for. There was poor visibility into how many applications they actually had, and no pre-existing enterprise architecture function.

No Visibility Into 1,100 Applications, Unchecked Acquisition Growth, and No Link to Business Capabilities

Like many health systems that have grown rapidly through acquisition, Ochsner Health had accumulated a technology landscape that had outpaced any single view of it. There were different finance systems, purchasing systems, licenses, and orphaned applications acquired in transactions. These all carried costs and complexity that no single team had visibility into.

Aaron describes this kind of cost as “parasitic,” a term that landed with his clinicians: spend that quietly eats away at the organization's finances without delivering proportional value.

For a nonprofit health system where, as he puts it, every penny matters, that was a problem with a direct line to patient care.

Underneath the cost problem was a language problem. Clinicians, physicians, and IS staff weren't speaking the same language about technology, which made decisions at scale challenging. There was no shared framework for explaining why a given application created or destroyed value, and no agreed picture of what the organization actually owned.

Building a Foundation, One Quick Win at a Time

Aaron had a long-standing relationship with Slalom's Managing Director of Technology Strategy & Advisory, M. Hans Delly, and trusted them to run a current-state assessment and, for the first time in Ochsner Health's history, build an enterprise architecture capability. Slalom recommended Ardoq as the platform to anchor that work. For Aaron, the deciding factor was the team's conviction in the tool, not a feature checklist.

“I looked in their eyes and saw their belief in Ardoq, and that was enough for me. The foundation was the trust I had in Slalom, and they proved themselves right.”
- Aaron Buck, CTO at Ochsner Health

The work started with capture. Slalom and the CTO office used Ardoq to inventory an application landscape that had never been documented. They then layered Business Capability Modeling over it to clarify what each application was actually for, which line of business it served, and who owned it on both the technical and business sides. The goal was a comprehensive view of what existed, how it was used, and what business value it created or eroded.

Crucially, Slalom's team built the capability model to be holistic rather than IT-specific, so it spoke to the chief medical officer as readily as to the IS team. As Hans puts it:

“Enterprise architecture here was treated as a business discipline, not a technology function, with the model providing clinicians and technologists with a common taxonomy and a shared language grounded in Ochsne's business.”
- M. Hans Delly, Managing Director of Technology Strategy & Advisory at Slalom

Reaching the people who held the knowledge meant going wide, not deep. Rather than interview application owners one by one across an organization of that size, the team used Ardoq’s surveys to engage roughly 700 application owners in 20 weeks.

Hans is candid that this took real change management: executive sponsorship from Aaron to set the context, clear communication about why the ask mattered, close tracking of completion rates, reminders, and quick responses when things went wrong. The survey data flowed back into Ardoq, correlated with capabilities, building an end-to-end view from value streams to capabilities to business processes, with the application landscape and financial data overlaid. Some of that financial data required, in Hans's words, “spreadsheet gymnastics,” pulled from CMDBs, ServiceNow, and spreadsheets alike.

The output stakeholders carried into rooms was Ardoq Presentations, segmented by audience: executive-level views for senior leaders and deeper breakdowns by org unit. When people asked the natural questions, where is this coming from, and can I trust it, the team could take them from the presentation back to the live source of truth in Ardoq.

Key Outcomes With Ardoq

1) $16M in savings from rationalization reinvested in transformation

The headline outcome is financial. Working through the landscape, the team identified a savings target north of $16 million, a significant number for a nonprofit health system and, as Aaron notes, an indicator of how much maturity the organization had been missing.

2) 1,100 applications fully enriched with owners, costs, and business as well as technical fit

Those savings came from surfacing redundant and duplicative applications and providing leaders with the evidence to make hard consolidation decisions.

3) Aligned tech portfolio to business capabilities in a shared language

Just as important is the shift in how decisions get made. With a shared capability model and a live source of truth, Aaron and his team can now sit down with clinicians and show, in terms they relate to, where effort and capabilities are being duplicated. That reframes difficult conversations. Aaron was candid with a room of physicians early on that he would eventually be "taking away your shiny toys," the applications they'd been promised, because as a nonprofit, every penny has to have optimal impact. The capability model gave that message a factual basis.

4) Established a framework for AI standards and governance

For Aaron, the savings are the fuel, not the finish line. With their newfound visibility, Ochsner Health's in-flight transformation program is set to deliver an AI-ready landscape that will help them accelerate the responsible adoption of AI in patient care.

Aaron's advice to healthcare system CIOs and CTOs

"You need enterprise architecture in place with a platform like Ardoq because it helps IT leaders contain costs, drive value, and provide a way to be better advocates for the business."
- Aaron Buck, CTO at Ochsner Health

Parasitic costs are going to eat away at your finances. Applying "intelligent architecture" will be a key capability for your organization to ensure you're achieving the intended impact without introducing additional cost. AI compounds the risk if you don't have the architecture to support it, and a clear capability picture enables intelligent conversations with clinicians about what to adopt and what to sunset. A data-driven EA approach is also vital to ensuring that the organization's application ecosystem doesn't gradually become costly and bloated all over again.

Looking Forward

The foundation is built to enable what comes next. The clearest near-term move is modernization: Ochsner Health is on a path from on-premise data centers to the cloud, with Slalom, Ardoq, and AWS as partners in that transformation. As Hans describes it, the two existing data centers weren't performing as the organization believed, and the destination was the cloud from the outset, both to migrate existing applications and to set the foundation for modernizing or retiring them. Enterprise architecture work is a prerequisite: you can't migrate, modernize, or retire what you can't see.

Beyond the migration, Aaron and Hans describe applying the same architecture discipline to Ochsner Health's AI ambitions, an approach they call “intelligent architecture”: aligning AI capabilities to business capabilities so the organization can deploy AI in a consistent, governed way rather than, as Hans puts it, the “wild, wild west.” With AI focused first on the clinical side in one of the most underserved regions in the country for health services, Aaron ties it back to the mission plainly: better, faster tools for clinicians help save lives. 

The Gap Between AI Ambition and AI Readiness Isn’t Unique to Ochsner

What Ochsner is working through isn't unique to Ochsner, and that's rather the point.

Hans has worked with enough organizations to spot the Ochsner pattern. A new technology leader steps into the role, often a CIO, CTO, or now a chief AI officer, and has to do three hard things at once: get their arms around the business, establish credibility with their peers, and figure out what they actually have to work with. Every one of those is a point of friction, and all of them affect a leader's success.

Into that pressure comes the board's AI ambition. In Hans's experience, most organizations respond by jumping in with both feet. What set Aaron Buck apart was the willingness to say the unpopular thing: not yet. Hans explains that the AI journey isn't linear; every organization is at a different point on it, and the precursors have to be right first. That message rarely lands well in a room of executives who have already decided AI is the priority. Aaron made the case anyway that, without the architectural foundation, AI deployment is guesswork, and you can't understand the implications of the changes AI introduces across the organization.

For Hans, the first unlock is never the technology. It's language. When people say "AI," he points out, they're saying two letters that everyone in the room hears differently, filtered through their own experience, so a group can discuss the same topic while talking straight past each other. The job is to get senior leaders aligned on what they're actually trying to achieve, at Ochsner, elevating the quality of care across the Gulf Coast, and then on the gap between that ambition and the organization's current ability to execute. Naming that gap honestly, and agreeing on how to bridge it, is where the real work starts.

That reframe extends to enterprise architecture itself. Hans describes a pendulum that swings across his profession's history: EA is treated as critical, then dismissed as an ivory-tower function that only says no, then rediscovered, again and again. He and Ardoq have aligned on breaking the cycle by reframing EA as a business discipline rather than a technology function, something both technologists and business stakeholders have to own. That also means redefining the architect.

A modern enterprise architect, in his words, is bilingual: able to sit with the chief revenue officer, the chief medical officer, or the CFO and understand the business objectives they're driving, then translate those into what the technology organization has to deliver. The value of EA, he argues, is in that translation, bridging the two conversations rather than living in one of them.

It's a useful frame for any leader facing the same board pressure Aaron did. The advice underneath it is simple: get the foundation right, get everyone speaking the same language, and treat architecture as the thing that makes ambition executable, not the thing that slows it down.

About Ochsner Health

Ochsner Health is the leading nonprofit healthcare provider in Louisiana, Mississippi, and across the Gulf South. Founded in New Orleans in 1942, it operates 47 hospitals and more than 370 health and urgent care centers. In 2024, more than 40,000 team members and 4,900 employed and affiliated physicians cared for 1.6 million people from every U.S. state and 63 countries.