For more than a decade, IT departments have been quickly increasing their spending on cloud computing to hundreds of billions of dollars annually. That’s led some tech leaders, including Akamai Technologies chief technology officer Robert Blumofe, to rethink their cloud strategies.
“Our approach has evolved considerably, and in ways that are fairly consistent from what I hear from our customers, some of whom are Fortune 500 [companies],” says Blumofe.
Blumofe sees three major themes related to tweaking his approach to the cloud: the hunt for cost savings, new thinking on the design and architecture of cloud-based applications, and the emergence of generative artificial intelligence, which relies on the cloud’s data storage to train and run large language models.
Cost savings has been a particularly big focus at Akamai, which helps customers with streaming and cybersecurity. Blumofe has been seeking relationships with alternative third-party software companies beyond the public cloud hyperscalers, such as Amazon Web Services, Microsoft Azure, and Google. He wants to free Akamai from vendor lock-in with the cloud giants, which often bundle their services with proprietary software.
Much of this pivot to disentangle software and cloud is from “Project Cirrus,” an initiative that Blumofe has led to migrate some application workloads to Akamai’s cloud hosting service Linode, which the company acquired for $900 million in 2022.
“We save a ton of money because we are just using the core infrastructure as a service that is much more reasonably priced,” says Blumofe. He estimates he has cut spending on public cloud providers by 40% in the first year of Project Cirrus.
Worldwide annual spending on public cloud services is projected to increase 21% to $723.4 billion in 2025, according to research firm Gartner. With so much money at play, there are plenty of opportunities for CIOs and CTOs to overspend.
“You really need to optimize the whole stack to be able to drive efficiency and performance,” says Ranjit Bawa, U.S. chief strategy and technology officer at consultancy Deloitte. Bawa advises clients to rethink their approach to every aspect of the cloud, including compute, storage, and even the physical data center facilities. Switching vendors, including to emerging cloud players, could help cut costs.
Jim Swanson, chief information officer at pharmaceutical giant Johnson & Johnson, says with cloud computing needs constantly rising, he leans on an internal operational framework known as FinOps, which brings together IT, finance, and business division leaders to align on how best to tap the cloud at an effective price.
“We’re constantly looking at how we buy [cloud] at the best cost possible,” says Swanson
J&J’s multi-year cloud migration is still under way, with the company continuing to modernize systems that support functions like supply chain and finance. A disproportionate amount of workloads are with AWS, though Swanson has embraced a multi-cloud approach that lets him match the right cloud service with J&J’s desired business outcomes, like when the company runs applications to help assist with simulations for orthopedic devices or develop new drugs.
“Given that we are a multi-sector company, and have such a big geographic footprint, you need optionality,” says Swanson.
Cybersecurity provider Palo Alto Networks says as more Fortune 500 companies increase their spending on cloud computing, there’s also a corresponding fear of losing control.
“You are losing control,” says Nir Zuk, CTO and founder of Palo Alto. “However, the cloud offers an opportunity to be much more secure compared to traditional data centers.” Zuk says by moving away from legacy physical data centers, companies can take a fresh look at developing cloud architecture guidelines that can more securely store and run their data.
With more customers embracing a multi-cloud approach, Palo Alto’s pitch is that its cybersecurity offerings are built to run across all cloud environments seamlessly. “It makes their lives much, much easier,” says Zuk.
C3 AI, meanwhile, in November cemented a partnership with Microsoft that integrates some sales and marketing efforts between the two companies. Under the terms of the five-year deal, Microsoft’s sales team will be able to pitch around 130 of C3 AI’s software applications to joint customers. The sales team is incentivized by making commission on any sale of C3 AI’s offerings.
Thomas Siebel, CEO of C3 AI, says he considered multiple vendors, but that Microsoft’s expertise selling enterprise software aligned most neatly with C3 AI’s business priorities. The deal also makes Azure the preferred cloud provider for C3 AI, tightening a relationship that began in 2018.
“They made it very clear they wanted to go to business now, they wanted to go fast, and they wanted to go to business globally,” says Siebel.
John Kell
Send thoughts or suggestions to CIO Intelligence here.
NEWS PACKETS
Chinese AI startup DeepSeek jolts the AI industry. America’s dominant AI industry was jostled after Chinese AI startup DeepSeek unveiled an AI model that rivals leading models like OpenAI’s o1 but claims to cost less to develop and require fewer Nvidia chips. U.S. tech stocks tumbled on Monday, with Nvidia hit especially hard, as investors rethought the global competitive landscape and the sky-high valuations that U.S.-based companies have commanded due to their strong position on AI up to this point. China’s breakout moment came with some mixed moments: DeepSeek’s AI Assistant usurped ChatGPT as the most-downloaded free app in the U.S. on Apple’s App Store, but the company also had to temporarily limit registrations because of a cyberattack.
OpenAI unveils ‘Operator’ Agent. Last week, just before the DeepSeek news cascaded across the globe, OpenAI announced that its ‘Operator’ agents had gone live for some users, creating the ability for AI to automate tasks like ordering groceries or booking travel. The new feature, which is available as a preview to ChatGPT Pro users (cost: $200 a month), is the latest advancement in the increasingly competitive world of AI agents. AI hyperscalers have christened 2025 as the “year of AI agents” and business software giants including Microsoft and Salesforce have been developing and releasing agentic AI tools that perform routine work tasks including summarizing reports and helping with a sales pitch.
UnitedHealth says 190 million Americans were impacted by a 2024 cyberattack. UnitedHealth nearly doubled the estimated total number of individuals impacted by the February 2024 cyberattack on the company’s subsidiary Change Healthcare, making it the largest healthcare data breach ever in the United States. The prior estimate was of around 100 million people and the final number is due to be confirmed with the U.S. Department of Health and Human Services. The data that was stolen includes names and addresses, Social Security numbers, passport numbers and medical information including diagnoses, medications, and test results.
AI hyperscalers are pouring billions into data centers. Meta has estimated it will spend $60 billion to $65 billion this year on capital expenditures, up sharply from the initial goal of up to $40 billion. Much of the money will go toward building and expanding data centers, which provide the compute power for Meta’s AI products. Similarly, OpenAI recently touted a $500 billion plan to build data centers in the U.S., with an upfront commitment of $100 billion from OpenAI, Softbank, Oracle, and Emirati investment firm MGX. But as Fortune reports, AI remains a big gamble with very little clarity into how much CIOs and other technologists are willing to pay for AI tools. For every success like cloud computing there are big flops, like Meta’s costly and unsuccessful bet on the metaverse.
ADOPTION CURVE
Data leaders warn there isn’t enough reflection after gen AI pilots fail. A survey of 600 data leaders conducted by Wakefield Research on behalf of cloud data management company Informatica found that 67% of companies have been unable to advance even half of their generative AI pilots to production. The stalled or abandoned projects have been affected by issues with data readiness and a lack of AI fluency and data literacy among employees. The average success rate, including companies that haven’t gotten any pilots into production, is just over 38%.
Data leaders say they continue to face pressure from the C-suite to increase spending on generative AI projects and that there's a strong desire to produce quick investment returns. But rather than reflect and adjust their strategy on AI before moving forward with new projects, 92% of data leaders are “concerned” that new generative AI pilots are moving forward without addressing the underlying problems that caused prior initiatives to fail.
JOBS RADAR
Hiring:
- Anne Arundel Dermatology is seeking a CIO, based in Linthicum Heights, MD. Posted salary: $190K/year.
- McDonald’s is seeking a senior manager of technology deployment, based in Chicago. Posted salary range: $149.3K-$190.3K/year.
- Kohler is seeking a head of IT for the company’s hospitality business, based in Kohler, Wash. Posted salary range: $183.6K-$239.2K/year.
- Constellation is seeking a director of IT, based in Nashville. Posted salary range: $198.9K-$221K/year.
Hired:
- The Office of the Federal CIO hired Gregory Barbaccia as its next leader, a political appointment under the Trump administration that does not require Senate confirmation. He will oversee federal spending on technology, IT policy and spending, and IT systems. Barbaccia previously served as CISO at Theorem and spent 10 years in leadership roles at Palantir Technologies.
- Marcus & Millichap appointed Evan Wayne as CIO of the commercial real estate brokerage, where he will oversee technology to support the company’s corporate team, clients, and brokers. Previously, Wayne was CIO for World Market and VP of technology at Rodan + Fields.
- Foundation AI announced Tom Majchrowski as CTO, joining the AI company that focuses on legal documentation, after he most recently served as VP of engineering at healthcare software provider Reveleer. Prior to that, Majchrowski was CIO at Dynamic Healthcare Systems and as VP of IT at National General Insurance.
- Phoenix Motor named Tony Zhou as CTO, effective January 20, where he will spearhead the development of the company’s electric buses and trucks. Before joining, Zhou was a head of advanced software engineering at electric vehicle manufacturer Lucid Motors and also held leadership roles at auto makers General Motors and Chinese state-owned FAW.
- Oncourse Home Solutions named Arul Sugunakumar as the first-ever CTO at the provider of home warranty and home repair plans, overseeing all internal and external technology functions. Sugunakumar joins Oncourse after most recently serving as CTO at Pampered Chef. He also held leadership roles at Accenture and Aon Hewitt.
- AIT Worldwide Logistics announced Ann Nemphos has moved to the role of CIO, after initially joining the shipping services company as CTO in May 2024. Nemphos will retain her CTO responsibilities in her new role and will report directly to chairman and CEO Vaughn Moore. Nemphos was previously CTO of World Group and held technology leadership roles at The Hartford and Travelers Insurance.
- MetaPhase Consulting named Brian Funk as CTO, joining the IT and management consultancy to oversee product development, software engineering, and cloud architecture. Most recently, Funk was a partner at IBM and also held leadership roles at Octo and NCI.
- Prometheus Hyperscale announced John Gross as CTO, after serving as a lead engineering consultant at the data center company for nearly four years. As CTO, he will oversee the engineering team and work with key partners to develop the company’s flagship 1 megawatt project in Wyoming. Gross has served as a principal mechanical engineer for clients including ExxonMobil, HP, and Valero Energy.
- Robin AI appointed Tramale Turner as CTO to oversee the legal AI-focused company’s software product development. Turner was previously CTO of TaxBit, head of engineering at Stripe, and held leadership roles at Nintendo and Volkswagen.