When Danaher bought the biopharma business that became Cytiva in 2020, the business changed hands before its data changed homes. The systems still lived at GE. According to Centroid’s account, keeping them there meant fees approaching $5 million a year. The purchase agreement had created a technology problem with a running meter.
- Centroid moves and manages enterprise systems, with Oracle at its core.
- Its customers need migrations that respect deadlines, dependencies, and daily operations.
- The useful lesson: design the intermediate step as carefully as the destination.
The assignment suited Centroid, a technology services company founded in 1997 by Scott Morrell, Eric Reed, and Jim Brull. It began as an Oracle consultancy. Today, the work stretches from cloud migrations and database management to financial planning applications, software development, and AI. Its territory is the machinery behind the business: systems whose importance becomes obvious when they stop.
A useful detour
Cytiva’s move happened in stages. Databases first went to existing Danaher hardware while the application layer moved to Oracle Cloud. Later, the databases followed into Exadata Cloud Service. That temporary arrangement accelerated the exit from GE. Centroid reports a nine-and-a-half-month migration, more than $6 million saved, and another $1 million a year saved on hardware and real estate.
Read the sequence closely. The final destination did not have to be ready before the expensive dependency could be removed. An intermediate state had economic value. For anyone planning a separation or migration, that is a useful question to carry into the first architecture meeting: which constraint needs to disappear first?
The decision behind the diagram
Subaru’s case offers a different lesson. In Centroid’s telling, the automaker initially intended a hybrid cloud approach for its Starlink system. Two things changed the decision: public cloud could shorten development work, and infrastructure capacity could expand with demand. The architecture followed those needs. The name of the architecture came second.
That distinction matters to a buyer. A migration proposal earns its keep by explaining a business constraint in technical terms. Faster provisioning matters when teams are waiting for environments. Scalable infrastructure matters when demand is growing. A fashionable destination, by itself, tells a finance director very little about whether a move is worth doing.
“It should be a partnership and you have to work together.”James Cornell / Cytiva
Somebody has to answer on Tuesday
Centroid’s commercial model extends beyond getting workloads across the line. It sells ongoing operations, support renewals, cloud capacity and consumption, and consulting. Its three managed service tiers make the division of labor unusually visible. Essentials covers foundational support. Elevate supplies specialist resources under customer-led project management. Elite lets Centroid take responsibility across more of the technology stack.
This gives an IT leader a more precise purchasing question than “Do we outsource?” A team might keep control of its roadmap while buying database or DevOps expertise. Another might want a provider managing infrastructure and applications together. The useful comparison is between responsibilities: who monitors, who fixes, who approves, and who carries the work into the following week?
Oracle is the center, with room around it
Centroid sits among specialist enterprise consultancies and managed service providers. A buyer could also consider a large systems integrator, such as Accenture or IBM, or build the capability internally. Centroid’s case rests on concentrated Oracle experience across databases, infrastructure, and business applications. The company reports more than 15,500 cloud workload migrations, a measure of accumulated work rather than a guarantee for the next project.
Its acquisitions show where it has wanted more depth. EPMI added Oracle enterprise performance management expertise in 2023. Intelletive brought DevOps and software development capabilities in 2024. Qbit expanded application expertise and Latin American reach that same year. For customers, the potential benefit is fewer gaps between application knowledge, engineering, and ongoing support.
The Oracle focus also travels. In April 2025, Centroid announced Google Cloud Partner status covering consulting, migration, managed services, and resale support for Oracle workloads on Google Cloud. In September, it announced managed security services with Stellar Cyber. Both additions address practical questions that survive a migration: where should the workload run, and who watches it?
AI still has prerequisites
EBS VisionIQ takes the same installed-system logic into generative AI. Centroid presents it as a way to use Oracle E-Business Suite data for insights and automation across functions such as finance and operations. The company advertises deployment in under a week. That is a vendor claim whose usefulness depends on the starting environment.
The compatibility list supplies the necessary detail: EBS 12.2.7 or later, Oracle Database 19c, Full Use licensing, APEX, and specified OCI services. An older or differently licensed installation needs assessment before that promise becomes a plan. The transferable habit is simple: ask for the prerequisites alongside the demonstration. They tell you how much preparation the attractive part requires.
The recurring business of change
In July 2026, Gridiron Capital announced an investment in Centroid as VSS exited. VSS said revenue and EBITDA had tripled during its five-year holding period, which included four acquisitions. Gridiron’s announcement emphasized recurring work and plans for further growth. There is a straightforward attraction here: enterprise systems need attention long after a migration project ends.
For an organization with little Oracle software, the specialist argument is weaker. A small team with straightforward hosting needs may also need less service than Centroid’s enterprise portfolio offers. The fit begins with the installed systems and the work required to keep them useful.
The buyer’s discipline remains the same. Price the workload, define the support boundary, and compare the proposed move with the cost of staying put. Cytiva’s temporary bridge offers the clearest image of Centroid’s work. Sometimes the valuable engineering decision is the step that lets a business leave one place before everything is ready in the next.