A trio of cloud industry leaders have launched a new company with the mission of modernizing software configuration data management.
Emerging from stealth today with $4 million in funding, ConfigHub is the handiwork of CEO Alexis Richardson, founder of cloud-native container management platform Weaveworks; CTO Brian Grant, former Google software engineer and original lead architect of Kubernetes; and CPO Jesper Joergensen, who led in various product roles at Salesforce (including Heroku) before joining Twilio to lead its voice, video and platform teams.
Richardson told TechCrunch that their new company is setting out to “get people out of config hell.”
If this just made you blurt out, “huh?” let us explain. In July last year, CrowdStrike issued an awry configuration update to its Falcon Sensor security software, causing widespread havoc and more than $5 billion in losses for Fortune 500 companies. Delta Airlines said it lost $500 million as a result of flight disruptions, kickstarting what will be a lengthy legal battle to recoup its losses.
That event demonstrated that software is now critical infrastructure, held together with a web of interdependent components connected by APIs to create powerful systems and applications. However, if any part of that web becomes compromised, deliberately or otherwise, it can bring down not just the whole house, but the entire town.
This brings us to configuration data, which ensures all these disparate parts play nicely with each other and that the software and system understand how they’re supposed to behave. But these configuration files can be an unwieldy mess, and ConfigHub is setting out to address just that.
“The problem is that configuration data is scattered all over the place — it’s become a total sprawl,” Richardson said.
Fixing DevOps
Much has changed in the way software is deployed and managed over the past 30 years. In the days of yore, enterprise software was mostly likely deployed via CD-ROMs on localized hardware, and configuration was limited to a few text tiles that instructed the operating system or applications where to find the things they needed to run properly.
Today, it’s not quite so straightforward. Gargantuan configuration files with thousands of lines of code is required to make the software run in what are often dynamic, large-scale environments.
Joergensen says he learned a valuable lesson at Heroku (a Salesforce-owned platform-as-a-service): Code and configuration data are not the same, and require very different approaches. Configuration data defines system settings and behaviors that can’t be debugged in the same way as software code can. But understanding this data — whether with tools like TerraForm or Kubernetes — is vital for avoiding costly misconfigurations that can cause outages or delays.
“We have made great progress on evolving collaborative coding practices with tools like GitHub,” Joergensen said. “But the configuration of live infrastructure needs a different approach. It is not just a bunch of files. Our goal is to bring the elegant app developer experience, pioneered by Heroku, to every kind of live production operations.”
As such, ConfigHub promises to “unify configuration management with modern, automated development workflows and compliance.” Instead of having to go hunting for the right piece of configuration that corresponds to a given error, on ConfigHub, everything will be held in a single database, making configs easy to find, replete with a live view that shows what the system is actually doing.
“So if your customer can’t access the system, you should be able to see and experience what they are seeing and experiencing,” Richardson added. “The configuration data in our system will be an expression of the live production system, which allows the team not only to identify the data, but also to change it, so that they can fix the customer’s problem immediately.”
The product, which will be served via a SaaS model, will focus initially on Kubernetes DevOps tooling such as Helm, Argo, Flux, Terraform and its open source fork Opentofu. Richardson says that ConfigHub is already working with some “medium to large” enterprise design partners, though he did not reveal names.
The product is still some months from formal launch, so today’s announcement is more about ConfigHub introducing itself to the world and announcing its $4 million investment. The funding comes from some notable VC and angel investors, including Crane Venture Partners, Encoded Ventures, Pear VC, and Poolside CEO Jason Warner.
The AI factor
While the CrowdStrike episode was one of the most costly configuration clangers in recent times, it wasn’t an isolated incident by any stretch.
“Configuration changes have, for a long time, been known as the leading cause of cloud system failures,” Grant told TechCrunch.
Last year, for example, Australian superannuation fund UniSuper was hit with an outage that left some half-a-million members without access to their accounts, after a Google misconfiguration led to UniSuper’s Google Cloud VMware Engine (GCVE) private cloud being deleted.
And back in January, a configuration error on GitHub’s cloud meant all Git operations were unavailable for up to two hours.
The problem is going to worsen — AI will compound matters at a time when companies are racing to embed AI into the fabric of their software.
“AI completely upends how the entire world writes, makes, and operates software entirely, not just code writing,” Warner said in a statement. “The next few years are much more than just vibing about small changes to greenfield code bases. Enterprises need mature solutions to hard problems. ConfigHub is a missing piece to allow this rewrite to take place safely.”
Companies can, of course, already build their own solutions from a concoction of tools such as Terraform, Kubernetes, Prometheus and others. But these can be overly complex to set up and manage.
There are big-name incumbents, too, such as ServiceNow and Atlassian that offer services to address configuration issues. But such companies were founded more than 20 years ago, before cloud computing had really taken off. Moreover, many such companies have long-expanded into the realms of HR, CRMs, team collaboration and project management.
“Those kinds of tools are not well-suited to the modern stack of dynamic, cloud-native, AI-powered, containerized applications,” Richardson added. “You can’t have somebody trying to roll out AI if a simple misconfiguration means that the AI will see their company data and publish it on the internet.”
Three’s a cloud
Prior to ConfigHub, Richardson founded an enterprise-focused cloud messaging company called RabbitMQ, which was acquired by VMWare subsidiary SpringSource in 2010. Then in 2014, he founded Weaveworks, which went on to raise more than $60 million in funding from a who’s who of investors including Accel, Amazon Web Services and Google Ventures. At Weaveworks, Richardson also developed the GitOps framework, which is now widely used in cloud-native and Kubernetes environments for managing infrastructure and application deployment.
Weaveworks formally shut shop last February, the result of “lumpy” sales and lack of fresh capital, Richardson wrote at the time. So in mid-2024, he teamed up with Grant and Joergensen to start ConfigHub.
Grant and Richardson met through their work at the Cloud Native Computing Foundation (CNCF), which counted Kubernetes as its inaugural project after Google donated it in 2015. Around 2020, Grant started giving up his various leadership roles around the Kubernetes project, and over time, he developed an idea for a new business whose focus likely wouldn’t align with Google’s priorities. This would eventually align, though, with Weaveworks coming to an untimely end last year.
Making the jump from one of the planet’s biggest technology companies to the world of startups seems like a huge transition. However, Grant worked at two startups before Google — for one, he was chief architect at a small startup called PeakStream, which Google acquired in 2007.
“I was the third engineer there — when we started, we didn’t even have a name, we didn’t have an office, we didn’t have a business plan,” Grant said. “So I have done very early startups before.”
While Google continued to grow exponentially in his 17 years at the company, Grant’s involvement with Kubernetes, which Google open sourced in 2014, was also a little akin to working in a startup.
“Kubernetes started as just four engineers and a product manager; we didn’t even have a manager in common, all of us reported to different people,” Grant continued. “It was very much like a startup in its own right.”
Richardson, meanwhile, met Joergensen around the time VMware acquired RabbitMQ more than 10 years ago. After serving in various senior roles at some of the world’s biggest tech companies, Joergensen too has now elected to step into the world of startups.
“You can get so much more done now than ever before — we’re jumping right into the rip-current of AI, and I probably wouldn’t want to be any other place right now,” Joergensen said. “In a startup, we get to really see with clarity what can be done.”
