When Rich Aberman, General Partner at Open Core Ventures (OCV), calls himself a believer in "arranged marriages," he's not talking about romance—he's talking about founder partnerships. And nowhere is this philosophy more evident than in the co-founders driving Kedify, an autoscaling company built on the popular open source project KEDA.
At first glance, the pairing seems unlikely: Zbynek Roubalik, a Czech engineer who spent over a decade at Red Hat maintaining critical Kubernetes infrastructure, and Jeff Fernandez, a seasoned New Jersey entrepreneur with two decades in HR tech and zero experience in cloud infrastructure. When Zbynek Roubalik received an email from OCV about starting a company around the open source project KEDA, he was intrigued. When OCV later suggested a non-technical CEO co-founder from HR tech, his reaction was blunt: "Are you crazy?"
When Jeff was introduced to the idea of leading an enterprise Kubernetes autoscaling company, he was all-in. “At first, I was taken by the OCV model,” said Jeff. “Then I looked at the portfolio, and I was really excited to meet Kedify. Zbynek and I hit it off right away. It felt natural.”
A few months into their partnership, their "arranged marriage" is proving that the best founder relationships aren't always the most obvious. The co-founders build on each other’s strengths to cover different business areas. “Zbynek has a very cogent point of view on the product roadmap that he articulates well. He services customers and onboards them. He’s provided a really strong foundation for me,” said Jeff. “It’s a great fit because we are focused on growth, and that’s where Jeff is an expert,” said Zbynek. “It’s a great combination.”
OCV's approach challenges the conventional wisdom that founders must have deep domain expertise from day one. Rich Aberman draws from his own experience: "I did not grow up wanting to build payment infrastructure for vertical SaaS companies. Yet, I ended up running product for my startup, WePay, and had a successful exit." His philosophy: "I think knowledge is cheap, especially today. It's very learnable. I don't think there's any substitute for ambition and aptitude."
The "arranged marriage" metaphor is intentional. Like successful arranged marriages, the partnership is based on complementary strengths, shared values, and mutual commitment to success rather than pre-existing chemistry. As Zbynek puts it with characteristic pragmatism: "Titles are not important. I don't care about titles. The important thing is if you can deliver what you would like to deliver."
Two unconventional founder paths merge
Zbynek's path to entrepreneurship wasn't linear. Growing up in a small Czech village, he was torn between his love of technology and history. "I probably wouldn't be able to fund my lifestyle with history," he recalls, "so that's why I chose technology."
After university, he joined Red Hat as an intern, eventually becoming part of the team that created KEDA—the open source Kubernetes Event-driven Autoscaler that has become the de facto standard for autoscaling in cloud-native environments. Companies worldwide now rely on KEDA to automatically adjust their computing resources based on real-time demand, making it one of the most critical pieces of modern infrastructure.
But Zbynek found himself at a crossroads. "I was in a moment of my life when I was thinking, okay, I'm doing this engineering stuff for a long time. I would like to try something else. But at the same time, I was so attached to KEDA, I don't want to completely abandon this space."
When OCV reached out, the timing was perfect. Zbynek knew Sid Sijbrandij's reputation from GitLab and was intrigued by OCV's focus on open source companies. More importantly, he recognized a clear commercial opportunity: while KEDA serves the majority of use cases well, enterprise customers needed additional capabilities for mission-critical deployments, real-time autoscaling, and enterprise support.
The validation came quickly. Even before bringing on a commercial co-founder, Zbynek demonstrated remarkable business instincts by closing a six-digit contract—an extraordinary feat for a purely technical founder without sales experience.
Jeff's journey to Kedify was equally unconventional. After 20 years building HR tech companies—including growing Grovo to $20M ARR before selling to Cornerstone OnDemand—he was winding down his DEI analytics company, Dandi, when OCV's head of recruiting, Mel Condos, reached out.
"She's like, 'It's really technical. Your background is probably going to be a problem. Otherwise, I think it could be great,” said Jeff. "And I was like, you know, yes and yes, I'd like to meet Zbynek if you'll let me."
What attracted Jeff wasn't just the technology, but the market dynamics: "Buyers know exactly what they want. They typically have very clear problems, they have budget, and they're very direct. All of that was pretty exciting for me." Jeff’s attunement to the buyer and the buying process ultimately won over Zbynek. “When I started, there was so much I needed to learn. A lot of things around the business I didn’t know,” said Zbynek. "Jeff’s experience is helping us shape the business. He fills the gap that we had in the company."
Jeff embraced the learning curve while focusing on what he could immediately contribute: scaling go-to-market operations, business development, and operational infrastructure. Rather than forcing an artificial division of labor, they let responsibilities emerge organically around each founder's strengths.
“ You’ve [Zbynek] run the company very well,” said Jeff. “You service customers, you onboard them. It’s unlocked an opportunity for me to focus on growth and marketing at the top of the funnel and make the mid-funnel execution that much better.”
Instead of siloing each other into their respective areas, the two work collaboratively to navigate every aspect of the business. “We’re building a foundation of mutual trust,” said Zbynek. “I can handle the technical stuff, and I’ve learned a lot about sales, but it’s not enough and too much for one person. At the same time, Jeff joins customer calls, and he’s learning a lot about the product. We work together.”
Autoscaling for the AI boom
With Zbynek's profound technical expertise and Jeff's keen understanding of buyers and market dynamics, their combined strengths have significantly evolved Kedify's product and company vision, charting a clear path forward. This synergistic partnership has allowed Kedify to not only commercialize existing open source technology but also to develop sophisticated solutions that address critical enterprise challenges KEDA alone cannot solve, particularly as organizations move towards more complex, microservices-based architectures and the burgeoning AI infrastructure revolution.
The company has developed sophisticated HTTP-based autoscaling capabilities that go beyond traditional CPU-based metrics, enabling more responsive and intelligent resource management. "CPU specifically doesn't work well as a leading indicator," Jeff explains. "It is a lagging indicator of usage. It doesn't sense what's coming. Whereas Kedify has over 70 connectors and you can create your own metrics. What are the leading indicators specific to your business that predict that usage is coming?"
This technical differentiation has proven especially valuable as organizations move toward more complex, microservices-based architectures where traditional autoscaling approaches fall short. For enterprise customers running mission-critical workloads, the ability to scale proactively rather than reactively can mean the difference between seamless user experiences and costly outages.
Perhaps most significantly, Kedify has positioned itself at the center of the AI infrastructure revolution. As organizations increasingly want to run their own AI models rather than rely solely on external services, for compliance, cost, or performance reasons, they need sophisticated autoscaling for GPU-intensive workloads.
"AI workloads are even more sensitive to traffic than HTTP traffic," Zbynek notes. "When you are running AI workloads, scaling and serving them properly on Kubernetes is even more important because there are multiple different metrics that you need to watch, and GPUs are expensive."
The company is currently working with design partners to expand from CPU to GPU autoscaling, targeting the growing market of organizations running large language models and other AI workloads in their own infrastructure. Given the cost of GPU resources—often orders of magnitude more expensive than traditional compute—the potential savings from intelligent autoscaling are substantial.
Individual strengths build collective success
While it’s just the beginning for Kedify, the foundation of mutual respect, complementary skills, strong early traction, and shared ambition suggests this unlikely pairing is building something significant. In a world where technical founders often struggle with go-to-market execution and commercial founders can lack technical credibility, Kedify's success demonstrates that sometimes the best partnerships come from the most unexpected combinations, especially when there's a massive market opportunity waiting to be captured.
With strong customer traction, positioning at the intersection of Kubernetes adoption and AI infrastructure needs, and an ambitious growth trajectory, Kedify's story is just beginning. Their success proves that with the right combination of technical excellence, market opportunity, and founder chemistry, even the most unlikely partnerships can drive remarkable growth.