Tag: Cloud

Posted on

Cloud cost management startup CloudZero lands $32M investment

CloudZero, a cloud cost management startup, today announced that it raised $32 million in a Series B funding round led by Innovius Capital and Threshold Ventures.

Bringing the company’s total raised to over $52 million, the tranche will be put toward expanding CloudZero’s platform and scaling its go-to-market efforts, CEO Erik Peterson says — with an emphasis on savings insights and self-service analytics.

“The pandemic accelerated digital transformation for many companies, which increased software development and spend in the public cloud, while the broader slowdown in tech increased the focus on profitability and unit economics and drove a shift away from innovation at all costs,” Peterson told TechCrunch in an email interview. “CloudZero has benefited from both of these trends and with this recent fundraising is positioned to accelerate our growth as demand for our platform is at an all-time high.”

Peterson co-founded CloudZero with Matt Manger nearly ten years ago after identifying what he describes as an “intrinsic” connection between efficient architecture and cost-effective cloud solutions. The way Peterson sees it, every cloud engineering decision is essentially a buying one — yet the cost conversation often bypasses the engineers who drive these decisions.

To attempt to solve for this, CloudZero provides engineering teams with cost data in a unified dashboard, combining the data with different types of business- and system-level telemetry. All this is laid on top of cloud billing data, enabling CloudZero to answer questions such as how much an individual product or feature costs.

CloudZero’s tech relies on hourly cloud spend data to alert engineers of “abnormal” spend events (as identified by an AI algorithm). The platform ingests customers’ cloud, platform-as-service and software-as-a-service spend to normalize it into a common data model.

“Cloud adoption is now widespread across sectors, but digital businesses struggle with lack of visibility, control, and optimization of their cloud costs — which impacts their ability to manage cloud spending,” Peterson said. “With up-to-the-minute monitoring, our platform identifies cost drivers and anomalies instantly, enabling engineers to proactively address cost spikes or inefficiencies and avoid cloud waste.”

Whether CloudZero’s platform is actually “up-to-the-minute” and “instant” remains up for debate. But undeniably, there’s demand for cost-cutting cloud solutions.

According to a recent IDC survey, 55% of cloud buyers believe inflationary pressures over the past few months have negatively impacted their return on investment. In fact, those organizations claim that the cloud now makes up on average 32% of their IT budgets; a quarter of respondents say that they’re looking internally at where they can make cloud cost reductions.

CloudZero doesn’t stand alone in the cloud cost optimization market, also known as FinOps — which isn’t exactly surprising given the huge market opportunity. There’s ProsperOps, which closed a $72 million investment in February for its software that automatically optimizes cloud resources to deliver savings. Other players include Xonai, Vantage, Cast AI and Zesty, as well as larger vendors like VMware and Apptio.

For its part, CloudZero has over 100 customers. Peterson wouldn’t disclose revenue, but said that it’s increased 10x since 2021.

Cloud cost management startup CloudZero lands $32M investment by Kyle Wiggers originally published on TechCrunch

Posted on

Ampere launches a new cloud native processor family with its first custom-designed cores

Over the course of the last few years, thanks to its Altra and Altra Max processors, Ampere made a name for itself as one of the premier suppliers of Arm-based chips to large cloud services providers. These chips were also based on Arm-designed cores, though. Today, the company is officially launching the Ampere One family, its first set of processors with custom-designed chips highly optimized for data center use cases.

While the Altra Max chips, which started shipping in 2021 were based on Arm’s Neoverse N1 architecture, these new cores were designed in-house and built on a 5nm process.

“It is time for us as an industry to meet the moment and embrace change. Our future growth as an industry depends on it,” Ampere CEO Renee James said. “The cloud has ushered in a whole new world and approach to software development. Isn’t it time for the microprocessor to do the same?”

These processors remain compatible with the Arm instruction set, so there’s nothing new developers have to do to support them, but more importantly, by designing its own cores, Ampere was able to optimize these processors for its core user base of large cloud platforms. The new AmpereOne chips can support up to 192 cores, a significant upgrade from the Altra Max chips’ maximum of 128. In the AmpereOne family, every core now also gets 2MB of L2 private cache (up from 1MB) and the company added several features that ensure that every core gets access to the same amount of memory bandwidth (or not, if that’s what the user desires) and provide more fine-grained power management, for example. There are also features like process aging monitoring, which helps these vendors better manage the lifecycle of the processor and monitor it for issues, as well as advanced droop detection and security features like memory tagging and single-key memory encryption.

The idea here is to build a chip that is custom-designed for the cloud. “Not only are we delivering a processor, that is high performance, low power, but it’s also one that’s uniquely well suited for the cloud. That’s why we call it the cloud native processor. We built it from the ground up for this use case,” Ampere CPO Jeff Wittich told me.

Ampere prepares to launch its first custom data center chips

As Wittich told me, the company increasingly hears from customers who are looking to reduce the carbon emissions from their data centers and manage their energy costs — a problem that’s especially acute for data center operators in the European Union. Ampere — and Arm’s server-centric chips — have long promised to offer more performance per kilowatt than the competition from Intel and AMD. Ampere now mostly describes this as performance per rack and argues that in its benchmark of running a typical web application, a rack of 36 Ampere CPUs provides the same performance as two racks of AMD Epyc CPUs and three racks of Intel Xeon CPUs. The company didn’t provide updated benchmarks for the AmpereOne processors, but Wittich argued that its performance per watt is comparable to the Altra family.

“Being able to go in and say: I can provide you a server that consumes half the power and deliver twice the performance of the alternative that’s a pretty easy sell for [OEMs] now,” Wittich told me. “Obviously, some people now understand the sustainability angle to it, too.” And even if sustainability is only a second-order problem for some data center operators, he noted, they are now increasingly limited by the power available to them in some metropolitan areas — plus, it’s getting harder to get new data centers approved, so many operators have to figure out ways to increase the performance within their existing physical footprint.

Current Ampere customers include the likes of Google Cloud, Microsoft
Azure, Oracle Cloud, Alibaba, and Tencent, as well as OEMs like HPE and Supermicro. It’s easy to spot the one large cloud provider that is missing from this list, but AWS has long bet on its own Graviton line of Arm-based chips, so that’s not likely to change anytime soon. 

While Ampere will continue to offer its existing Altra family and has no plans to discontinue it, the company also won’t be launching any new chips in this range. “All of our future processors will rely on our own custom cores,” Wittich said. “The big thing about Ampere One that makes so impactful is that this is the start of every one of our future products. We’ve got our custom cores out there now and that’s the basis for everything.”

Ampere launches a new cloud native processor family with its first custom-designed cores by Frederic Lardinois originally published on TechCrunch

Posted on

AWS announces new version of Aurora database that strips out I/O costs

AWS announced the general availability of a new version of its Aurora database this week called Amazon Aurora I/O-Optimized. The big news from this version is that it gets rid of all I/O charges for database use, a move that should reduce overall database costs for customers with large workloads, and bring more predictability to their cloud database bills.

“With the new Aurora configuration, customers only pay for their database instances and storage consumption with no charges for I/O operations. Customers can now confidently predict costs for their most I/O-intensive workloads, regardless of I/O variability, helping to accelerate their decision to migrate more of their database workloads to AWS,” the company said in a statement.

And getting customers to migrate more workloads is the goal of course. But with more companies looking to operate more efficiently in the cloud, a product like this could appeal to increasingly cost-conscious CIOs.

But it’s important to note that it is a higher priced product than the standard Aurora database, according to Corey Quinn, chief cloud economist at The Duckbill Group, a consulting company that helps customers lower their AWS bills. “It’s an alternate pricing model. They charge more for this model as a baseline rate so it’s going to come down to the specifics of a given workload as to whether it’s a good idea to use it,” Quinn told TechCrunch.

In a blog post announcing the new version, AWS’s Channy Yun acknowledged that it depend on the type of workload. “You can now confidently predict costs for your most I/O-intensive workloads, with up to 40 percent cost savings when your I/O spend exceeds 25 percent of your current Aurora database spend. If you are using Reserved Instances, you will see even greater cost savings,” he wrote. As you can tell, and as Quinn points out, the devil will definitely be in the details of your particular workload requirements.

Ray Wang, founder and principal analyst at Constellation Research said it’s a win for customers with big workloads. “Normally every time you read data that’s not cached and then write data back to your mySQL or Postgres data, you incur an I/O charge,” he said. “This is designed to drop your pricing because they have found a more efficient way internally to handle this, and they’ve passed on the cost savings to customers as we enter an age of AI.”

This should be particularly helpful for customers with data-intensive workloads like AI or seasonal eCommerce use cases. Customers can bring new workloads or move between the standard Aurora database and the I/O optimized version in the management console, based on expected workloads, to help manage costs.

AWS announces new version of Aurora database that strips out I/O costs by Ron Miller originally published on TechCrunch

Posted on

Mavenir raises $100M to build more AI into its cloud-based automated network technology for telecoms services

One of the bigger trends in telecoms has been the evolution of cloud computing and how software operating in the cloud is being used by carriers to build out and manage their networks — moves disrupting traditional business models based around the purchasing and deployment of costly telecoms routers, switches and other costly physical equipment. Today, Mavenir — one of the hopefuls in that space, with a focus on Open Radio Access Networks and related services for 5G networks — is announcing $100 million in funding to expand its technology and its customer base.

The first of those will involve bringing more AI tools into the mix, the company said; and as for the latter, the company already has a sizable business: it currently has more than 250 service provider and enterprise customers across 120 countries, services that altogether cover more than half of the world’s mobile subscribers. The plan will be not just to continue expanding that pool, but to build more services to sell to current customers.

Mavenir described the funding as “anchored” by Siris, a PE firm that’s been backing the company for years. Mavenir is not naming further investors in this latest capital injection, but other existing backers include Koch and Pendrell. Koch and Siri also participated in the company’s last financing, a $155 million round in October 2022. It’s not disclosing its valuation but we have contacted the company to ask.

Mavenir has been through a couple of different ownership structures, including at one point being majority owned by Siris and used as the umbrella for a roll-up of telecoms software vendors Mitel and Xura — both also of the bigger consolidation trend — to capitalize on new opportunities with 5G networks. After selling a $500 million stake to Koch in 2021, Siris remained at least a partial shareholder, and given the $255 million raised in the last seven months, it seems Siris is back into investment mode with the company now.

Mavenir is part of the wave of companies that also includes companies like DriveNets, Arrcus, and even AWS looking to capitalize on the so-called trend for digital transformation in the telecoms industry. Telecoms carriers are on the hunt for more ways to bring down their own capital and operational expenditures while providing more flexible services to business and service provider customers as they scale up and down. Kit from the likes of Cisco, Juniper and Arista represents the legacy, incumbent approach to network provision and operation; now, the pitch is that software, delivered via cloud services, is the new and better solution.

“This new capital will allow us to accelerate our capabilities in automation, sustainability, and use of AI as we enable our customers to efficiently deploy and operate Open RAN based end-to-end cloud-native networks,” said Mavenir CEO and president Pardeep Kohli in a statement. “Our unique strategy incorporates best practices from the hyperscale, cloud and IT industries, to transform how the world connects and builds the future of networks.” Kohli has been with Mavenir for years — initially directly with the company and then reappointed the head of the business after his next employer — Xura, where he was also the CEO — was acquired and rolled up into Mavenir.

“This investment enables Mavenir to further scale its business and maintain its leadership in Open RAN and 5G transformation,” said Hubert de Pesquidoux, a Siris executive partner who is also chairman of Mavenir, in a statement. “We firmly believe in the automated networks of the future that are cloud-native, AI-native and Green-native, and we are confident that Mavenir’s innovations are essential in driving that evolution.”

Mavenir raises $100M to build more AI into its cloud-based automated network technology for telecoms services by Ingrid Lunden originally published on TechCrunch

Posted on

Cloud infrastructure revenue growth dips to 19% in Q1, but still hits $63B for quarter

As expected, companies are continuing to cut back on their cloud spend where possible, as tech budgets continue to get careful scrutiny. That resulted in a quarter in which the cloud grew 19%, up $10 billion over last year to $63 billion. That doesn’t sound too terrible, except when you compare the numbers to last year when the market grew 32%.

Clearly we are still in a cost cutting cycle and it’s having an impact on cloud infrastructure revenue growth, but Synergy Research reports that there are signs that we could be coming out of the recent doldrums. Synergy’s chief analyst John Dinsdale says that overall the market remains solid, and we are starting to see a shift in some of the trends that have been contributing to the the downward growth.

“There has been some angst about declining cloud growth rates, but the Q1 worldwide market value grew by more than $10 billion compared with the first quarter of 2022. Clearly the relatively weak economy has caused some enterprises to more closely review spending on cloud services, but the market continues to grow despite those challenges,” Dinsdale wrote in a comment to the press.

He points out that the Chinese market has returned to growth and exchange rate pressure has started easing, contributing to growth in the EMEA and APAC regions. “The law of large numbers pretty much dictates that growth rates must decline, but in absolute terms the market continues to grow at a healthy rate, driven by the fundamental benefits of cloud adoption,” Dinsdale said.

It’s fair to say that most industries would be happy with a growth rate that’s nearly 20% in this economic climate, but the cloud has been dealing with rates in the high 30s until recently, so it feels worse, and as we’ve learned, perception counts.

This is especially true for Amazon, where AWS has been the growth engine for the company for more than a decade and it’s suddenly looking at a quarter where the growth plunged into the teens to 16%. Again, for a mature company, that doesn’t feel that horrible, but the cloud revenue numbers continue to trend down from the 20% rate the company saw the prior quarter.

Meanwhile Azure growth also continued to drop too. While Microsoft’s cloud arm grew at a higher rate than Amazon at 27%, that was down from 31% the prior quarter. Google Cloud was up 27.5%, slowing from 32% the prior quarter, but turning a profit for the first time.

Is there really a march from the public cloud back on-prem?

What impact does all of that have on market share? Well, as it turns out, not that much. Amazon has owned a fairly steady third of the market for years, even as the pie has grown. Microsoft has been gaining slowly but steadily, and Google has reached 10% and so far is holding steady there. The big three account for 65% of the total revenue.

For the quarter, AWS continues to have 32% market share, good for over $20 billion for the quarter, Microsoft held steady from the prior quarter at 23%, good for over $14 billion in revenue for the quarter and with 10% Google Cloud took in over $6 billion.

Synergy looks at infrastructure and platform as a service, as well as hosted private clouds to come up with its market numbers.

Nothing goes up forever, but there are signals on the horizon that perhaps the cloud infrastructure market will return to growth. There is certainly still plenty of room, especially with these companies looking at adding data-intensive AI workloads, and with that, the market should stabilize over time.

Cloud infrastructure revenue growth dips to 19% in Q1, but still hits $63B for quarter by Ron Miller originally published on TechCrunch

Posted on

UK regulators could be right about cloud portability obstacles

U.K. communications watchdog group Ofcon reported last week that it was investigating cloud infrastructure vendors — paying particular attention to Amazon and Microsoft — for making it too difficult to move workloads from public clouds. This raised a legitimate question about the obstacles to portability.

The report pointed to three things in particular that Ofcon would investigate: egress fees, the payments these companies charge when customers want to move data from their platforms; general restrictions on interoperability and portability; and discounts they use to keep companies with large workloads on their platforms.

Ofcon is investigating whether cloud vendors, especially the largest ones, have been deliberately putting up roadblocks to keep customers from changing vendors, giving consumers fewer options once they’ve committed to a particular seller. That could put smaller competitors at a distinct disadvantage.

Ofcon typically looks at consumer issues like the cost of broadband, but it sees cloud computing as a public utility, where pricing has a direct impact on U.K. businesses. The inquiry suggests the U.K. authorities see a possibly deliberate attempt on the part of these companies to keep their customers in the fold.

Of course, every company wants to keep its customers from churning. That in itself is not necessarily a problem, but if these companies are setting up systems to make it difficult for customers to switch, that becomes an issue for groups like Ofcon. It’s worth noting that the United States Federal Trade Commission also announced an inquiry into public cloud vendors last month, asking for public comment on market power and security risks. It joins the EU, which launched an investigation into Microsoft last year.

When we asked Microsoft and Amazon about the report, they both said they are working with Ofcon and are committed to a competitive market in the U.K. Yes, that’s all well and good and exactly what you would expect from the vendors, but do these groups investigating anticompetitive behavior have a point?

UK regulators could be right about cloud portability obstacles by Ron Miller originally published on TechCrunch

Posted on

Britive, which helps secure public clouds, lands $20.5M investment

Proving that there’s still plenty of venture money in cybersecurity, cloud identity security platform Britive today announced that it raised $20.5 million in a Series B funding round. Led by Pelion Venture Partners with participation from Liberty Global Ventures, Crosslink Capital and One Way Ventures, the new brings Britive’s total raised to $36 million.

CEO Art Poghosyan says that it’ll be put toward expanding the company’s investments in customer successes, marketing and sales, as well as product development.

“Ironically, the pandemic affected Britive’s business more positively than not,” Poghosyan told TechCrunch in an email interview. “It accelerated the move to virtualizing work environments and transition from data center to cloud.”

There’s truth to that. According to a 2022 survey from Equinix, 71% of IT decision-makers plan to move more functions to the cloud over the next 12 months. The cloud has plenty of upsides, among them scalability and accessibility. But there are drawbacks, too — one being security (potentially). A poll from the Cloud Security Alliance reveals that close to one-third of businesses using a public cloud provider experienced security issues within the past two years, including unauthorized apps and poorly configured APIs.

Poghosyan says that Britive was built to address those cloud security concerns — particularly on the identity and access management (IAM) side. Founded in 2018, the startup’s platform generates access privileges — including tokens and keys — on demand for people and software accessing cloud services and apps.

“Our tech automatically expires and removes privileges when not in use to ensure the cloud keys and tokens are not exposed to attacks around the clock,” Poghosyan said. “Britive’s offering allows development teams to continue building at cloud speed while security teams maintain full visibility and control over cloud identities and privileges.

Britive is far from the only player in the vast and growing IAM market, though. There’s well-capitalized ventures like Saviynt, whose platform enables companies to secure apps, data and infrastructure in a single platform. Private equity firm Thoma Bravo paid billions to acquire identity access management startups SailPoint, Ping Identity and ForgeRock. On the less grandiose end of the spectrum, there are ventures like ConductorOne, which aims to bring automation to identity and access management.

Poghosyan asserts that Britive’s anomaly-spotting AI differentiates it from the others out there. Using AI, Britive analyzes cloud user activity logs in near-real time and attempts to identify behavioral patterns that could be dangerous or unauthorized. The platform then alerts the relevant security teams, giving them the option to quickly terminate any potentially problematic sessions.

TechCrunch can’t speak to Britive’s efficacy. But the company’s growth suggests there’s something there. Poghosyan says that annual recurring revenue is in the “millions” of dollars and growing 3x annually, and that Britive now has “dozens” of customers, including several Fortune 500 brands.

“Our platform manages thousands of daily active privileged identities who interact with critical business applications and infrastructure hosted on major cloud platforms,” Poghosyan added. “Even in a broader climate of economic slowdown, enterprise organizations continue to invest in future-proofing their cloud infrastructure and security solutions stack, which Britive is an essential part of.”

Britive, which currently has about 55 employees, expects to add 10 to 20 staffers by the end of the year, Poghosyan said.

Britive, which helps secure public clouds, lands $20.5M investment by Kyle Wiggers originally published on TechCrunch

Posted on

Nvidia partners with Google Cloud to launch AI-focused hardware instances

In partnership with Google, Nvidia today launched a new cloud hardware offering, the L4 platform, optimized to run video-focused applications.

Available in private preview on Google Cloud through Google’s G2 virtual machines, Nvidia says that the L4 platform is designed to accelerate “AI-powered” video performance. Serving as a general-purpose GPU, L4 delivers video decoding as well as transcoding and video streaming capabilities.

Beyond providing access to the L4 platform through Google Cloud, Google is integrating L4 into Vertex AI, its managed machine learning service for enterprise customers.

For those who prefer not to sign up with Google Cloud, L4 will be available later this year from Nvidia’s network hardware partners including Asus, Cisco, Dell, Hewlett Packard Enterprise and Lenovo.

L4 sits alongside the other AI-focused hardware solutions Nvidia announced today, including L40, H100 NVL and Grace Hopper for Recommendation Models. L40 is optimized for graphics and AI-enabled 2D, video and 3D image generation, while H100 NVL supports deploying large language models such as ChatGPT. (As the name implies, Grace Hopper for Recommendation Models is recommendation model-focused.)

L40 is available this week through Nvidia’s aforementioned hardware partners. Nvidia expects Grace Hopper and the H100 NVL, meanwhile, will ship in the second half of the year.

In related news, today marks the launch of Nvidia’s DGX Cloud platform, which gives companies access to infrastructure and software to train models for generative and other forms of AI. Announced earlier this year, DGX Cloud lets enterprises rent clusters of Nvidia hardware on a monthly basis — starting at an eye-watering $36,999 per instance per month.

Each instance of DGX Cloud features eight NVIDIA H100 or A100 80GB Tensor Core GPUs for a total of 640GB of GPU memory per node, paired with storage. With DGX Cloud subscriptions, customers also get access to AI Enterprise, Nvidia’s software layer containing AI frameworks, pretrained models and “accelerated” data science libraries.

Nvidia says that it’s partnering with “leading” cloud service providers to host DGX Cloud infrastructure, starting with Oracle Cloud Infrastructure. Microsoft Azure is expected to begin hosting DGX Cloud next fiscal quarter, and the service will soon expand to Google Cloud.

Nvidia’s aggressive push into AI compute comes as the company moves away from unprofitable investments in other areas, like gaming and professional virtualization. Nvidia’s last earnings report showed its data center business, which includes chips for AI, continued to grow (to $3.62 billion), suggested that Nvidia could continue to benefit from the generative AI boom.

Nvidia partners with Google Cloud to launch AI-focused hardware instances by Kyle Wiggers originally published on TechCrunch

Posted on

Nvidia launches new cloud services and partnerships to train generative AI

At its annual GPU Technology Conference, Nvidia announced a set of cloud services designed to help businesses build and run generative AI models trained on custom data and created for “domain-specific tasks,” like writing ad copy.

Under the brand Nvidia AI Foundations, each individual cloud service — Nvidia NeMo for language models and Nvidia Picasso for image, video and 3D content generation — includes pretrained models, frameworks for data processing, APIs and support from Nvidia engineering staff. Once models are ready for deployment, enterprises can run run them in Nvidia’s cloud or elsewhere.

As of today, the NeMo generative AI cloud service is in early access. The Picasso service is in private preview. Signing up for either requires submitting an application.

“Generative AI is driving the fast adoption of AI and reinventing countless industries,” Nvidia founder and CEO Jensen Huang said in a press release. “Nvidia AI Foundations let enterprises customize foundation models with their own data to generate humanity’s most valuable resources — intelligence and creativity.”

With the NeMo cloud service, Nvidia says that developers can add domain-specific knowledge and functional skills to text-generating models for use cases like chatbots, enterprise search and customer service. Models available on the service will be regularly updated with additional training data, the company says, and benefit from information retrieval capabilities that augment them with real-time data.

As for Nvidia Picasso, it’s a cloud service for building text-to-image, text-to-video and text-to-3D object models. (While text-to-video and text-to-3D object models are less common now than their text-to-image brethren, vendors are increasingly beginning to experiment with them — anticipating a surge in interest.)  Customers can use Picasso to train and deploy models on their data, either using pretrained models from Nvidia or optimizing and running their own models.

In a big customer win, Nvidia says that it’s working with Adobe to co-develop “the next generation of commercially viable generative AI models” on the Picasso platform. They’ll be brought to market through Photoshop, Premiere Pro and After Effects, according to Nvidia, filling out Adobe’s recently-debuted Firefly family of generative AI products.

No doubt eager to avoiding wading into the copyright controversy around generative AI, Nvidia says that it’s collaborating with another party, Getty Images, to train “responsible” generative text-to-image and text-to-video models. The models will be developed using content from Getty Images’ library, Nvidia says, with Getty Images will providing royalties to artists on any revenues generated from the models.

That’s significant given that Getty Images is currently embroiled in a lawsuit against Stability AI, maker of another generative AI service called Stable Diffusion, over using its images to train its AI without permission from Getty or rightsholders. Nvidia, it seems, didn’t want to take its chances — or perhaps wanted to hedge its bets, particularly as similar lawsuits work their way through the courts.

In a related announcement, Nvidia says it plans to partner with Shutterstock, a Getty Images competitor, specifically to train a generative text-to-3D object model using Picasso. Shutterstock’s assets and metadata will be used for training, and Shutterstock will compensate artists via its recently launched Contributor Fund, Nvidia says.

Shutterstock plans to offer the model on its own platform once it’s ready for deployment, complementing the text-to-image tool it already provides through a partnership and licensing agreement with OpenAI.

Nvidia launches new cloud services and partnerships to train generative AI by Kyle Wiggers originally published on TechCrunch

Posted on

Cast AI raises $20M to help companies reduce cloud spend

Cloud costs remain a top concern for organizations. According to a recent Anodot survey, 50% of businesses are struggling to control them, in part because they lack visibility into their cloud usage. Unsurprisingly, reducing those costs has become a top priority. A report from Wanclouds finds that 81% of IT leaders have been directed by their C-suite to reduce or take on no additional cloud spending.

Surely that’s music to the ears of startups whose tech is designed to reduce cost spend. There’s several out there. But one of the more successful vendors is Cast AI, which today announced that it raised $20 million in an investment led by Creandum with participation from unnamed existing investors.

Founded by Yuri Frayman, Leon Kuperman and Laurent Gil in 2019, Cast AI — powered by AI, as the name implies — analyzes a company’s cloud usage to attempt to find the optimal cost-performance ratio. Cast AI customers can connect their cloud plans across AWS, Google Cloud and Azure to see recommendations and have the system implement them automatically.

“We decided to build a platform that would help companies automatically optimize and reduce their cloud costs — without manual intervention,” Frayman, who serves as Cast AI’s CEO, told TechCrunch in an email interview. “The need for such a platform became clear after we co-founded Zenedge, a cloud-based cybersecurity platform that was ultimately acquired by Oracle. As our application scaled up, we watched our cloud bill exponentially increase from thousands of dollars to millions of dollars … Our goal [with Cast AI] was to build the product we wished we had at Zenedge.”

According to Frayman, Cast leverages “many models” to drive its cloud usage optimization automation engine, which scales cloud resources up and down in real time while optimizing for cost. The models are trained on usage metadata from customers, public cloud pricing and inventory information, and undisclosed “other signals” that cloud providers make available. (Frayman notes that customers who don’t wish to submit their data for training can request that it be deleted.)

Cast AI

Cast AI’s optimization dashboard, showing real-time data across public cloud instances. Image Credits: Cast AI

Some of Cast’s models are responsible for determining how often public cloud jobs might be interrupted, while others predict spare compute capacity availability. Still others are trained to anticipate seasonal changes in workload requirements, like decreases in cloud usage during the winter months.

“We’re able to predict lower future compute prices to impact future batch workload scheduling, kind of like searching for a cheaper flight on Kayak and booking a future date that’s cheaper,” Frayman explained.

As we’ve written about before, several years ago, the market for cloud optimization software, while nascent, was consolidating as incumbents in adjacent sectors saw an opportunity to make a mark. In 2017, Microsoft acquired Cloudyn, which provided tools to analyze and forecast cloud spending. VMware and NetApp bought CloudHealth and Spot (formerly Spotinst), respectively, within the span of a few years. Somewhere in the midst of all that, Apptio snatched up cloud spending management vendor Cloudability, while Intel purchased Granulate for $650 million.

The consolidation isn’t necessarily over. But some vendors are prospering, judging by their successful fundraises. For example, ProsperOps, a Cast rival, landed $72 million in a venture round that closed in February.

To stay competitive, Cast recently launched a “zero-cost,” trial-like cloud cost monitoring and reporting product and its first cybersecurity offering: a tool that shows users security best practice violations as measured by the Center for Internet Security benchmarks and shows ways to fix those violations.

“Our goal is to have fewer people doing higher-order, more interesting work in cloud operations,” Frayman said. “The industry is faced with a challenged customer that needs to save cloud costs immediately and cannot afford to parse through reports and assign people to solve the problem manually.”

Frayman declined to disclose Cast’s revenue numbers. But he did say that Cast’s quarter-by-quarter revenue has grown by over 220% since the platform launched. Thousands of customers’ apps are being optimized by the platform, he claims, and Cast is actively managing over a million public cloud CPUs.

He thinks the Silicon Valley Bank scare could spur business, too, given the newfound tightness of capital. To that end — in light of the bank failures — Cast is offering onboarding services for free for a limited time. Customers will pay only once they start saving money.

Frayman says that Cast will put the newly secured funds, which bring the startup’s total raised to more than $38 million, toward growing its team and “working toward being cash-flow positive.”

“We currently have 75 employees and expect to grow the team as the business scales, and as we continue to hit our metrics driven milestones,” he added. “We will add automated security capabilities into the Cast AI platform. We will also invest in other features to help our customers with ‘day-two’ operations, like monitoring, testing and alerting. That way, we can help maintain the performance and reliability of our customers’ clusters with the ultimate goal of minimizing or completely eliminating human intervention.”

Cast AI raises $20M to help companies reduce cloud spend by Kyle Wiggers originally published on TechCrunch