From Hypervisor to Next Generation Cloud: Uma Chingunde on Cloudflare TV

Uma Chingunde’s profile in our company directory says that if she were an animal, she’d be “a cat, because they live the best lives, nine times over.” And indeed, Uma is tracing the nine lives of a cat with her career. She chose tech over biotech to keep her options open as an undergrad in India; she joined VMware in the early days of virtualization to build what became an industry-leading hypervisor; she led Stripe’s Compute group which managed all cloud operations for an exponentially-scaling payments stack; and last year, she switched gears to become VP of Engineering at Render. I learned more about her career journey through two decades of technological advancement in a recent interview she did with Cloudflare COO Michelle Zatlyn on her show 'Yes We Can.' By switching roles and becoming a customer of virtualization and cloud technology at Stripe, Uma gained a unique perspective on both sides of the technological advancement she had contributed to. After becoming an expert in high performance computing products that appealed to both developers and enterprises, Uma proved she could also successfully manage compute infrastructure for one of the fastest-growing companies of all time, and scale a team from 12 to 40+ engineers to operate it. By bringing this experience to her leadership at Render, Uma is helping to turn her success at Stripe into a general success story for our customers. Her interview left me thinking about what makes a tech startup special enough to take a chance on. In this post, with Uma’s help, I’ll look at several special companies to understand what they have in common.

The power of abstraction

In a her interview on Cloudflare TV, Uma discusses Render’s opportunity to continue the huge advancement of technology her career has already spanned. Uma calls cloud computing “the defining part” of that advancement. While she acknowledges the sheer ambition of saying "We are going to build a new cloud provider” (fair — ambitious, and exciting!), Uma's career itself tells the story of why Render’s mission matters. Virtualization, pioneered by IBM and later VMware, seeded the potential for cloud computing, “and the ability to provide compute for everyone around the world was built on top of that.” Uma notes that the next generation of SaaS technology companies, Stripe for example, built their success from that ability by leveraging infrastructure-as-a-service. Why did Uma join Render? “Abstraction is a theme that I believe in. The most powerful companies are providing something that abstracts away complexity,” she observes. Stripe is a perfect example of this. Why should companies take the risk to develop a payment processing solution internally when they can offload that risk, and the complexity, to another tech company, for which doing the job well is the core mission?

The benefits of abstraction may seem obvious to us now, but in a bygone era of online transaction processing, companies often perceived more risk in taking the chance on an unproven startup like Stripe than in building secure payment systems themselves. But no company can forever ignore the valuable leverage of abstraction, and survive. Even VMware — already 4000 employees strong when Uma first joined as an intern in 2007 — was perceived as high-risk in the early 2000s. Sure, it had the potential to lower costs by dramatically increasing server utilization, but that came with a tradeoff of trusting a new and untested company.

Tech that’s worth the risk

As developers, it can be hard for us to identify which technology is worth taking risks on. While we shouldn’t jump at every opportunity to switch to something new and shiny, we can talk to stakeholders to understand whether a solution will allow the company to focus better on its core mission. When a piece of tech truly delights you by abstracting away time-consuming work, it’s okay to start a conversation about tool choice. Consider Docker. This implementation of OS-level virtualization borrowed technology born in the Linux community in 2001, but made it both highly accessible to developers, and extremely powerful in accomplishing their missions. The quality and usability of the Docker project and container format allowed Docker to enjoy early adoption by many technology companies well before it reached 1.0. This rare case paved the way for innovation in container orchestration, most notably Kubernetes, which enables us to automatically deploy Docker containers to perform at scale within Render and many other modern computing products. I classify Docker, like VMware, as “worth the risk” because it’s so good at abstraction:

• In packaging together application code and all dependencies, including OS-level ones, Docker allows apps to run reliably across computing environments.
• The abstraction enables engineers to stop thinking about compatibility across internal and external computing environments.
• The time spent thinking about dependencies and compatibility across environments can be re-allocated to companies’ core missions.

Regarding abstraction within the DevOps and infrastructure space, we can again learn from Uma. When discussing why she joined Render, she observes, “if there is an entire team at a large SaaS company that's only doing this thing, then likely there is a market there.” For her, ‘the thing’ is internal infrastructure for engineering; in leading Stripe’s Compute org, she helped solve problems for Stripe developers who needed “all of the complexity of the underlying cloud providers abstracted away, who just want to write their code and deploy it.” This abstraction powered Stripe’s key differentiator: velocity that enabled it to respond quickly to customer needs. This lesson is bread and butter for companies that have taken on the complexity of a specific problem and abstracted it away for customers:

• Stripe, which abstracts payments
• Twilio, which abstracts telephony communications
• Segment, which abstracts customer data governance

VMware also comes to mind, since they abstracted server utilization and management very successfully prior to the rise of cloud computing. Then AWS, and later GCP and Azure, operationalized the computing concepts pioneered by IBM, VMware, and the open source community, to completely abstract away the data center and offer it as a service. However, as Render CEO Anurag Goel, formerly Stripe’s Head of Risk, explains, gaining this “abstraction” costs companies dearly as they invest millions of engineering dollars to manage the growing complexity of AWS itself — essentially, to prevent application developers from spending their precious time on cloud operations and deployment — instead of enhancing the customer-facing products and services only engineers can build. It costs a lot to build your own developer experience. At Stripe Anurag and Uma, recognized this as having a high opportunity cost. Render, as part of a new generation of cloud providers, removes from the developer experience the complexity of DevOps and the first generation public cloud. We host all of developers’ services in one place and provide control and flexibility without the complexity of traditional IaaS. To help us do this well, Uma applies lessons learned throughout her career, including the importance of customer focus. “It’s almost taken for granted now that every company needs to listen to their users a lot more; the difference is companies that do this very well,” she says. Render subscribes to this model of listening and allowing customers to influence our roadmap and user experience. We know we must be customer-obsessed to make our abstraction “worth the risk.” We also know that the velocity our customers gain from Render allows them to respond to their own customer feedback much more successfully.

The right people

When we make more personal decisions about technology companies — namely, which are “worth the risk” to join ourselves — people and relationships often light our paths. Uma reflects, “I have always really enjoyed my work and have been successful when I picked a place where it was the right people,” and she’s right — smart, humble, kind people make a job, and a company, a success. When I interviewed with Uma, our conversation made a big difference in my decision to take the leap and join Render. Beyond my recognition that Uma had already known of our CEO Anurag at Stripe and chose to work with him again, I sensed a kinship to Uma’s spirit. Watching her interview on Cloudflare TV’s “Yes We Can,” I got a hint to its origin. When asked why she loves the systems and infrastructure domain, Uma tells Michelle about being one of two female CS students in all her systems courses, operating systems, distributed systems etc, befriending the other student and joining forces on projects. Uma recalls, “in undergraduate and grad school, there was this subtext where women would choose the more fun, frontend, UI thing. But I had a bit of a stubborn streak where I would not do what I was told or pushed to do; so, if I’m being pushed in this direction, I’m going to do the opposite.” This “stubborn streak” is me in a nutshell, so I’m grateful to Uma for the validation. She talks about her love of solving the really hard, really fun problems, and I think about my choice to pursue a career in infrastructure. I basically thought, here’s something I understand but feel sort of bad at. People don’t expect me to be doing this kind of work, it’s really hard, and sometimes I fail. But I’m never bored, and when I do figure it out, I feel amazing. So far, it’s working out. I haven’t been bored in a decade, and now I’ve started a new adventure at Render, where I can work with incredible role models like Uma. Building the next generation of public cloud might be a moonshot, but as Michelle Zatlyn at Cloudflare would say, "Yes We Can!" Render is an ambitious, long-term journey and we’d love for you to join us. Visit our careers page or DM us on Twitter to embark!