The trouble at Grupo Porteo was wonderfully indifferent to the language of modern computing. During hurricane season, damage and a leak in its data center made the logistics company reconsider where its systems belonged. Its shipping information depended on IBM i. The weather had no opinion about IBM i. Water simply found a way in.
- Lightedge runs infrastructure and the services needed to keep it useful.
- Its specialty is joining IBM systems, private cloud, and public cloud.
- The right answer can include leaving equipment exactly where it is.
Lightedge’s account of the project describes a buyer looking for IBM expertise, responsiveness, and affordable hosting. Porteo selected a provider within a week and implemented the solution within the month. The attraction was practical: keep the logistics system working without managing the building around it. As president and CEO Paul Sarrapy put it, “If the system fails, everything fails.”
01 / The cloud needs somewhere to live
That is a useful entrance to Lightedge. The company provides colocation, private cloud, managed public cloud, security, and disaster recovery. Colocation means housing a customer’s equipment with power, cooling, and connectivity. Managed cloud adds responsibility for operating the environment. Those are different purchases, even when they appear on the same proposal.
The distinction matters to buyers whose applications already have jobs: tracking shipments, processing orders, supporting patients. An infrastructure change has to preserve those jobs. Lightedge occupies the space between a company operating everything itself and a company assembling infrastructure, cloud platforms, and support providers separately.
02 / An old system can be a valuable system
IBM Power is central to the proposition. IBM i and AIX applications require Power infrastructure; moving them is not simply a matter of picking an ordinary public-cloud instance. Lightedge supplies specialized hosting and management, with connections to hyperscale platforms. An existing application can keep doing its work while newer services develop around it.
Now consider Red Gold. The tomato processor’s critical applications ran on IBM i in rural Indiana, where connectivity limited the appeal of moving them to the cloud. Skilled administrators were also difficult to find. According to Lightedge’s case study, Red Gold retained its on-premises infrastructure and brought in outside management. Its reported results included better monitoring, disaster preparedness, and unchanged IT headcount.
Porteo moved out. Red Gold stayed put. The same provider could serve both because the scarce resource was sometimes hosting capacity and sometimes the people who understood the machine. For a buyer, that is a more useful distinction than the age of the application.
Connect, monitor, secure, recover.
03 / Buying the missing pieces
Lightedge began in Des Moines in April 1996 as Lighthouse Communications, providing network services and managed hosting. Its own history describes a later retreat from the networking business to concentrate on cloud, colocation, and compliance. That was a choice about focus: fewer kinds of services, more attention to infrastructure customers could not casually interrupt.
Acquisitions subsequently widened the offer. Anschutz Investment Company bought Lightedge in 2008; GI Partners became its owner in 2021. The 2024 Connectria agreement added IBM, AWS, and Azure experience to Lightedge’s existing colocation and private-cloud capabilities. It also proposed adding six data-center locations to a footprint then described as twelve facilities.
Read that expansion as an attempt to follow customers across their estate. A buyer with IBM applications and AWS services needs the connection between them to work. Consolidating providers can simplify coordination, although the buyer still has to establish which team owns each task.
04 / The bill includes the people
Lightedge’s business combines contracted infrastructure and ongoing managed services with projects such as migration. IBM Power Cloud offers granular consumption, including processor allocations in tenths of a core. The economic question is therefore broader than buying a server: it includes staffing, licenses, recovery capacity, and the operational work the customer retains.
Lightedge’s Shipyard case study reports a 35% reduction in AWS spending. That is a customer result, with no disclosed dollar baseline in the public account. It is useful evidence for asking about optimization, rather than a percentage to paste into another company’s budget. Hardware, service fees, and migration effort still need a scoped comparison.
Reported for The Shipyard in Lightedge’s case study. One customer outcome; individual economics vary.
Other cases reveal what the work actually involves. For CourtCall, Lightedge rebuilt AWS infrastructure and wrote GitLab automation for rolling updates. Luma’s migration introduced autoscaling and deployments across availability zones, followed by ongoing support. Renting capacity was only part of either assignment; changing how software reached users was part of the value.
“If the system fails, everything fails.”Paul Sarrapy · Grupo Porteo
Security is another part of that purchasing decision. Lightedge offers round-the-clock monitoring, detection and response, and audit assistance. For a team that cannot staff those functions internally, outside support can fill a real gap. The sensible buying exercise is to name the gap precisely: which alerts receive a response, which systems are covered, and what evidence the customer receives. A longer service list only helps when it matches the work someone actually needs done.
05 / Start with the thing that must survive
The 2026 updates extend that operational theme. In May, Lightedge added IBM Power Virtual Server support. In September, it described EverPure SafeMode: immutable snapshots retained for seven days on supported storage. The snapshots add a recovery layer alongside traditional backup. Their defined window and eligibility matter more than a broad promise of resilience.
There is a method here that readers can copy. Identify the application whose interruption hurts most. Establish its hardware, connectivity, and staffing constraints. Then decide where it should run and who should operate it. Red Gold demonstrates why a cloud move can be inappropriate when connectivity cannot support the workload. A provider’s security credentials also require checking the scope of the service being purchased.
Porteo’s leak supplied an unusually clear prompt for that conversation. Most companies get less theatrical warnings: an administrator leaving, a hardware renewal, a support ticket nobody owns. Lightedge’s opportunity lies in those moments, when keeping the business open suddenly becomes a very specific engineering assignment.