Prepared remarks
Thank you for standing by, and welcome to Rackspace's Second Quarter 2026 Earnings Conference Call. I would now like to hand the call over to Sagar Hebbar, Investor Relations. Please go ahead.
Thank you, and welcome to Rackspace Technology's Second Quarter 2026 Earnings Conference Call. I'm Sagar Hebbar, Head of Investor Relations. Joining me today are Gajen Kandiah, our Chief Executive Officer; and Mark Marino, our Chief Financial Officer. As a reminder, certain comments we make on this call will be forward-looking, including, without limitation, statements regarding our financial guidance and outlook, our enterprise AI deployment plans, capacity targets and timelines, expected capital expenditures, revenue per megawatt and margin assumptions, our financing plans, cash flow expectations, our business strategy and product roadmap as well as shifts in our business mix. These statements involve risks and uncertainties, which could cause actual results to differ materially. A discussion of these risks and uncertainties is included in the Risk Factors and Forward-Looking Statements sections of our most recent annual report on Form 10-K and subsequent quarterly reports on Form 10-Q filed with the SEC. Rackspace Technology assumes no obligation to update the information presented on the call, except as required by law. Our presentation includes certain non-GAAP financial measures and adjustments to these measures, which we believe provide useful information to our investors. In accordance with SEC rules, we have provided a reconciliation of these measures to their most directly comparable GAAP measures in the earnings press release and presentation, both of which are available on our Investor Relations website. I will now turn the call over to Gajen for an update on the business.
Thank you, Sagar. Good morning, everyone, and thank you for joining us. I want to start by reviewing our progress toward a clear strategic goal: becoming the accountable provider and operator of the full enterprise AI stack from core to cloud to edge. For more than 25 years, enterprises have trusted Rackspace to operate complex mission-critical infrastructure across private cloud, public cloud and data centers worldwide. As enterprise AI becomes operational infrastructure, the things that have always mattered most to our customers — governance, security, sovereignty, resilience and accountability — matter even more. McKinsey & Company estimates that global inferencing workloads will surpass training by the end of 2026 and represent two-thirds of all AI workloads by 2030. Rackspace has tens of thousands of customers across our installed base and many are starting to harness inference to run their businesses more effectively. The questions they are asking are sharpening around sovereign AI estates that run through a model-agnostic, vendor-neutral ecosystem, and we have listened carefully. Every AI interaction — an employee query, an agent evaluating a transaction, a hospital reading clinical data — pushes decision-making towards the edge, driven by sovereignty, latency and variable load. We don't see cloud versus edge. We see one integrated environment with workloads placed wherever latency, cost, security and criticality dictate. We believe trust will become one of enterprise AI's most valuable currencies. This is why we have been very deliberate in building the right partnerships. Our managed compute and inference platform backed by partners, including AMD, Dell, Palantir and Uniphore, gives customers a clear path to cost-efficient, controlled enterprise intelligence that scales with them. As we build out our AI infrastructure capabilities, I am excited to welcome Pranav Nambiar as SVP and GM for AI Infrastructure. Pranav brings over two decades of experience designing and building complex infrastructure systems with AWS, DigitalOcean, Google and Microsoft. Most recently, as Senior Vice President and General Manager of AI and Data Cloud at DigitalOcean, he spearheaded the company's strategic pivot into a premier AI Neo Cloud, paving the way for the company to be recognized as a unique Neo Cloud with full stack AI infrastructure and data management. His best-of-breed knowledge and execution skills honed with some of the world's most demanding companies is exactly what we need in order to build and scale this important new business. This quarter, we entered into a partnership with AMD and strengthened our partnership with Palantir as we build out our enterprise AI solutions. AMD brings the accelerated and differentiated computing platform, while Palantir brings platforms that connect AI with enterprise data and operational workflows while embedding security permissions and governance. Rackspace then brings the knowledge, infrastructure, migration, cloud and managed operational capabilities needed to run those platforms reliably in production across all regulated and nonregulated industries. Our forward-deployed engineers work in the customer environment, focusing on high-value use cases and remaining accountable beyond the initial implementation. The customer retains control of its data and operating context while Rackspace provides governance and accountability across the environment. Under our definitive agreement with AMD, we plan to deploy an initial footprint of 30 megawatts of AMD-based compute across Rackspace data centers in phases from late 2026 through 2028. The architecture incorporates AMD Instinct GPUs and EPYC CPUs, enabling us to match workloads with the appropriate compute while remaining accountable for performance and operations. We believe the combination of more efficient silicon, smaller domain-specific models and intelligent workload routing can materially improve the economics of enterprise AI while simultaneously mitigating exposure to a single model, whether it be for bare metal, inference as a service, fully managed enterprise inference or enterprise AI cloud. The attractive economics of our new growth vector bears repeating with the following illustrative example. The first deployment is expected to be nearly 2 megawatts targeted for completion by the end of 2026. Capital expenditures for the first deployment are expected to be approximately $75 million. Our goal is to ramp to cumulative capacity of 15 megawatts by the end of 2027 and a total of 30 megawatts of capacity by the end of 2028. We expect to average $15 million to $20 million in revenue per megawatt deployed with some variability based on CPU, GPU and customer mix. This range translates to $450 million to $600 million in annual revenue for the full 30-megawatt deployment. We expect EBITDA margins in enterprise AI to be in the 50% plus range. We are evaluating financing for a significant portion of the compute hardware through a combination of OEM financing, equipment financing and other asset-backed credit facilities with the financing collateralized by the newly acquired hardware. Early demand signals give us reason to be optimistic about the pace of deployment. Given the market's continued demand for high-performance compute and AI infrastructure and the long lead times for Greenfield and Brownfield data center projects, Rackspace is well positioned as we have the infrastructure, power, cooling and talent already available to us, and we have placed our initial order for AMD GPUs and CPUs. Alongside inbound calls, our optimism is also driven by our installed base of enterprise customers interested in adding capacity as well as co-selling by AMD, Palantir, Uniphore and our growing base of FTEs. On the platform side, we continue to see increased traction in our strategic relationship with Palantir, both in pipeline and signed deals. Across these engagements, we are seeing a consistent pattern in the type of problem customers bring us in to help solve: turning fragmented legacy data environments into unified AI-ready platforms and doing it fast with measurable ROI. Each deployment compounds what we have learned, making the next one faster and more repeatable. We will have more specifics to share as these engagements mature. Finally, our corporate focus continues to reflect where the market is heading and what our customers want. When we announced our One Rackspace initiative, our intent was to redirect the capabilities we have built over time to take advantage of the generational secular market opportunity in front of us. Enterprises are no longer choosing a single public or private environment for all their applications and data. Instead, they are asking for integrated architectures based on their requirements. As we move forward, we intend to communicate with you with that in mind to better represent our strategy and our milestones. Our priority will continue to be disciplined around capital and talent deployment as we move towards higher-yielding services and a strong balance sheet. And with that, let me get into our business performance, starting with Private Cloud. Second quarter Private Cloud revenue was $263 million, ahead of our July 9 guidance. The upside was driven by the timing of revenue recognition for a long-term customer contract. Excluding this impact, revenue would have been in line with our previously guided range. Because this recognition timing pulled forward revenue from future periods, including the second half of 2026, it does not change our full year guidance. We expect Private Cloud to grow this year even as we absorb supply-related timing impacts and strategically pivot the business towards higher-margin revenue. We will stay opportunistic about deals that accelerate our strategic pivot as they arise. Our customer wins this quarter reinforce a consistent story. Enterprises in regulated industries are choosing Rackspace to modernize and operate environments where governance, reliability and compliance are nonnegotiable as the foundation for AI adoption. For example, in health care, we deepened our relationship with AdventHealth, whose Epic EHR, one of the top five Epic systems in the world, we already host and manage. This quarter, that relationship expanded substantially. We signed a five-year agreement to host and manage infrastructure that lets AdventHealth greatly reduce their on-premises data center footprint and retire a separate disaster recovery colocation contract. This comes alongside a large-scale migration of roughly 366 applications, 2,300 virtual machines and 283 database servers onto Rackspace hosted infrastructure with full DR failover. We also added a new nonproduction Epic environment to support their IT development pipeline. Epic Managed Services is proprietary Rackspace IP, purpose-built for the governance and uptime clinical environments require. This is exactly the foundation regulated health care organizations need as they move AI from experimentation into production. In financial services, we strengthened our position in cyber resilience with a top U.K. banking firm. We signed a multiyear agreement to deploy and manage a first-of-its-kind cyber recovery cloud built on Rubrik alongside managed backup and managed Kubernetes services, supporting their next-generation development and test banking platform. This is the first phase of what we expect to be a multiphase deployment extending into staging and production. Finally, I want to share where our software strategy in private cloud stands as a critical part of Rackspace's ability to stitch the full stack together. We recently completed the production release of RackAI, our inference and fine-tuning platform that lets customers integrate AI into their workloads and applications through a simple API. Looking ahead, we'll continue building out RackAI with additional enterprise capabilities, including intelligent model routing and access to customized model harnesses, giving customers more flexibility as they scale their AI initiatives. Now for our Public Cloud update. Public Cloud revenues were $407 million. Public Cloud continues its pivot towards higher-value services-led work. We are aligning our capabilities from cloud adoption through AI in production, concentrating investment in the data and AI-led enterprise transformation, AIOps-driven managed services and forward-deployed engineering talent operating across cloud, core and edge. This quarter's wins reinforce our role as a trusted partner in regulated mission-critical environments. In the Americas, we were selected for a competitively bid federal defense engagement, building a multi-cloud management practice with FinOps automation and self-service capabilities. We also expanded a multi-work stream engagement with a major U.S. commercial airline, modernizing its cloud platform and embedding AI-powered development across its engineering organization to improve observability and systems availability. In EMEA, we deepened our relationship with a U.K. financial services organization, expanding into a full end-to-end managed services engagement and becoming their strategic partner on a multiyear modernization and AI road map. These wins reflect our strength in regulated data-intensive industries, deploying AI at scale while maintaining reliability, compliance and operational excellence. We also expanded our Public Cloud portfolio this quarter with a set of entry point offerings across clouds. These are the tips of the spear, structured often partner-funded engagements that open the door with a customer and expand into larger managed services relationships. Each one drives revenue for Rackspace and consumption for our partners, which is why AWS, Microsoft and others are funding them. The best example is our optimization and modernization assessment powered by AWS, a fully AWS-funded engagement that turns infrastructure and licensing optimization into a single business case for enterprises carrying heavy licensing obligations. The customer gets a funded road map, Rackspace earns the position to execute it and the workloads land on our partners' platform. In addition, we launched offerings on the same model this quarter across Microsoft Copilot adoption, managed network security and data readiness. The common thread across this quarter's launch is structured funded engagements that convert enterprise AI ambition into governed production-ready deployments and a clear path into Rackspace's broader managed services relationship. Our moves this quarter strengthen and expand our role as a trusted partner and operator alongside curated best-of-breed ecosystem partners, where we are accountable and where data sovereignty belongs to our customers and no one else. That is what today's Rackspace will continue to deliver. With that, I will turn it over to Mark for our financial results.
Thank you, Gajen. In the second quarter, total company GAAP revenue was $670 million, up 1% year-over-year, with the beat versus expectations primarily driven by the aforementioned timing impact within the Private Cloud segment. Non-GAAP gross margin was 18.6% of GAAP revenue, down 120 basis points year-over-year. Non-GAAP operating profit was $27 million, flat year-over-year as continued operating expense discipline, including payroll savings from workforce reductions, was partially offset by higher professional fees and marketing expenses. Non-GAAP loss per share was $0.08 compared to $0.06 in the prior year quarter. Cash flow from operations was negative $32 million and free cash flow was negative $48 million. We ended the quarter with $111 million in cash and $202 million in total liquidity, including the undrawn portion of our revolving credit facility. During the quarter, we opportunistically repurchased $11.2 million in aggregate principal of our senior notes pursuant to a previously established 10b5-1 plan. This reduces our go-forward interest expense and further strengthens our capital structure. We remain focused on deliberate, steady deleveraging while continuing to fund strategic growth priorities. First half free cash flow reflected known seasonal cash uses, including annual incentive compensation payouts, strategic vendor prepayments and debt repurchases. We expect free cash flow generation to accelerate through the second half, strengthening liquidity as the year progresses. Turning to segment results. Private Cloud GAAP revenue was $263 million, up 5% year-over-year and ahead of the $242 million to $246 million outlook we shared on July 9. The upside reflects a benefit from an embedded lease treated similarly to a hardware sale under a customer's managed hosting contract, a timing item tied to a long-term contract rather than a change in underlying volume. Non-GAAP gross margin was 32.3%, down 460 basis points year-over-year, driven by the same revenue item noted above, along with slightly higher customer licenses and data center costs. Non-GAAP segment operating margin was 21.8%, down approximately 280 basis points year-over-year. In Public Cloud, GAAP revenue was $407 million, down 2% year-over-year, reflecting lower infrastructure and services revenue. This was modestly ahead of the $399 million to $403 million outlook we provided on July 9 due to higher reported consumption from hyperscaler partners than anticipated. Non-GAAP gross margin was 9.7%, up approximately 10 basis points year-over-year on savings from workforce reductions. Non-GAAP segment operating margin was 4.7%, up 80 basis points year-over-year on improved operating expense efficiency. Turning to guidance. Consistent with our July 9 call, we expect full year GAAP revenue of $2.45 billion to $2.55 billion, a decline of 7% year-over-year at the midpoint. The vast majority of that change reflects our strategic decision to exit low-margin Public Cloud revenue over time. From a segment perspective, we expect Private Cloud revenue of $1.0 billion to $1.05 billion, up 4% year-over-year at the midpoint, and Public Cloud revenue of $1.45 billion to $1.5 billion, down 13% year-over-year at the midpoint. We expect total non-GAAP operating profit of $125 million to $135 million, up 3% at the midpoint. Adjusted EBITDA is expected to be $285 million to $295 million, up 5% at the midpoint, with non-GAAP loss per share of $0.25 to $0.30. Our non-GAAP tax rate is expected to be 26%, while non-GAAP other expenses will be in the $220 million to $230 million range. Non-GAAP share count is expected to be between 250 million and 260 million shares, excluding any dilution from the ATM program, as future issuance under the program will depend on prevailing share price and market conditions. Importantly, in the second half of 2026, we expect strong free cash flow and expect $50 million to $70 million in positive free cash flow for the full year. With that, I will now turn it back over to Gajen for final remarks.
For all businesses, especially those in regulated industries, the generational shift to the era of AI is imperative and no different than the shift to client-server networks in the '90s, the Internet in the early 2000s and mobile in the 2010s. Over 25 years, Rackspace has earned the trust of tens of thousands of customers as the accountable operator guiding their business through these infrastructure transitions. With the introduction of Rackspace's managed compute and inference platform, we will continue to be that accountable partner, provider and operator of a secure, reliable, flexible and controlled full enterprise AI stack. That is Rackspace. Thank you to our customers, partners and every Racker. With that, back to Sagar.
Thank you, Gajen. We will now go ahead and open the line for any questions. If you have any follow-up questions after today's call, please reach out directly at ir@rackspace.com. Operator, please go ahead and open the line for Q&A.
Questions and answers
Our first question comes from the line of Bradley Clark of BMO.
I want to ask about the Palantir business. It seems like the partnership has been off to a strong start. And just wondering if you could comment on your vision of how this partnership evolves over time and where you see it specifically adding to Rackspace's growth profile over the next several years.
Thank you. This is Gajen. I appreciate the question. Yes. The Palantir partnership over the last four to five months has truly gained traction, especially as enterprise customers have started to understand the importance of what I would call institutional sovereignty in terms of where models run, where the data needs to be kept sovereign versus what data can stay private versus public. As that wave has taken off from a commercial perspective for Palantir, we are also seeing a similar trajectory with our customers. What's been interesting is the type of opportunities we are doing follow a process: boot camp to agent camp to the first production use case and then scaling from there. A few things that have stood out to me are that with our Palantir engagements we land small and then scale fast. The boot camps and agent camps give us a good basis of understanding for what our customers want to do, and once you get that first production use case in, you're able to scale from there. The Palantir platform allows us to do that very quickly. The type of opportunities we are seeing are cutting across industries — renewable energy, packaging, specialty chemicals, and health care are some of the early wins we are seeing. Applying the start-small, scale-fast model is generating strong momentum. This has given both Palantir and Rackspace the opportunity to come together and structure a strict and fast go-to-market motion, which we believe will yield significant revenue. We're starting to see the early stages of that, and I'm really excited about the partnership as well as the type of customers and opportunities that we are building in the pipeline.
Our next question comes from the line of David Paige of RBC Capital Markets.
Maybe just a little bit of follow-up to that last question. It sounds like you're in a good position to start and continue to win regulated industry business. I was just curious if you could flesh out some of the demand that you're seeing from both existing customers and potentially new customers out in 2027. Just how do you see the customer base evolving? And what's the current demand and competitive environment for these regulated AI production environments that you seem to be the leader in?
Thank you, David. Great question. Coming into this, the orientation was really all about our enterprise customers and the regulated markets in which we operate. As we have expanded the foundational partner footprint with AMD, Palantir, Uniphore and others, we are seeing that partner motion add to both the opportunities we are seeing and the type of work we are able to do. A third piece that gives me significant confidence is the emerging trend around sovereign AI and the concept of sovereignty going beyond country boundaries down to enterprise boundaries, therefore needing modularity around the types of models you use, the type of compute you use, the orchestration required and the different types of inference that are starting to emerge. For instance, we do context-aware inference to go from model to model without losing the context of a particular request or an agent requirement. All of that ultimately means data needs to land somewhere that is safe, secure and governed. I think that's what gives us this extreme confidence in terms of the focus and strategic pivot we have made, and I'm seeing that market emerge very quickly. You would have seen recently that other companies are starting to really talk about this space emerging quickly as well. Regulated was where we started, but it feels to me like almost all enterprises will look for some form of a regulated approach — not necessarily regulated in terms of formal regulation, but an approach that provides a high degree of confidence in terms of where the data resides, how it gets processed and how it gets applied. That puts us in a very strong position.
Great. That's very helpful. If I could sneak one more in. You mentioned balancing growth investing and debt repayment and cash flow generation. So I was wondering if you could just provide a little bit more color on how you see that balance evolving over the next 12 to 18 months.
Yes. This is Mark. Good question. It's something we're obviously thinking about and it's high on our list. With the GPU investment itself with AMD, we expect a lot of that financing to come via hardware-backed or asset-backed facilities, so not necessarily impacting the cash we need to run the business. We're being judicious about making those investments by optimizing or restructuring other parts of the business to free up funds to drive those investments while continuing to grow operations and EBITDA. We believe we have significant growth levers with very high ROI that warrant investment over the next few quarters, including the GPU investment and the Palantir initiative. It's top of mind as we think about where to allocate our dollars while we continue to drive liquidity to free up for other investment initiatives.
As there are no further questions in queue, that does conclude the Q&A session and our conference for today. Thank you for participating. You may now disconnect.