Partha Seetala, 46, CEO of Robin.io, started as Robin Systems, a company he founded in 2015 and acts as the CTO at the same time. Before that, he worked at Veritas then Symantec always in storage products groups. He also spent some time at Kazeon and was CTO of a stealth startup. Outside of buildings products and teams, he is passionate about films and their story behind, how they are made, from idea, concept to production and the thinking of the director.
StorageNewsletter.com: Robin.io is a bit confidential as a company even if it’s getting better recently. Give us a snapshot of the company with financial, funding, employees, offices… with founders background as well.
Seetala: Robin.io is a company headquartered in San Jose, USA. The team is made of strong technologists with domain expertise with a track record of building mission critical software used at the core of our economy today. We have built a strong technology differentiation with 65+ patents in the areas of cloud-native storage, networking automation and orchestration. The company has been doubling its revenue in each of the last 3 fiscal years.
How much did you raise and do you need a new round?
We have raised $67 million so far and coupled with the revenue growth we have been seeing we did not have the need to raise additional funding. Every fast-growth startup is always looking at the cap table and opportunities to dial in the optimal amount and types of funding. That said, our present funding is aligned with our growth and product development plans.
As a pure software company, did the Covid-19 impact your business?
Like a lot of engineering-oriented companies, we had to scramble a year ago to establish new processes in the wake of Covid lockdowns, travel restrictions, and office closings. We ended up with new workflows and collaboration approaches that allowed us to meet design milestones and respond to customers while still respecting the upended personal lives that many of our employees faced because of pandemic restrictions. In the end, we have met our deliverables and launched important new products. We are incredibly proud of our team, and we are looking forward to our first opportunities to begin gathering in person again, which we can re-evaluate after winter. As for the market, we have experienced stronger than expected demand for both our enterprise and 5G product groups.
What were the reasons to launch Robin? Is it due to Kubernetes anticipated opportunities and DevOps storage needs?
Robin was launched based on the observation that every 5-6 years the tech industry disrupts the status quo by delivering a simplified technology stack to solve the use cases of that time. Notice that users interact with applications, and applications depend on infrastructure to meet the user’s need, however we observed that every project usually starts with elaborate planning at the infrastructure layer, which results in delayed roll-outs, dissatisfied users, and high cost. We wanted to change that. Our vision was to drive a true user+application oriented platform that made infrastructure invisible – users simply click a button to deploy an application and the infrastructure self-composes to meet the performance and availability needs of the application. When we launched Kubernetes was still in its infancy. But the rapid growth of Kubernetes and the industry consolidating around it as the de facto cloud-native platform accelerated Robin’s growth and helped us in realizing our vision.
What are the pains you solve and what are the associated uses cases you excel in?
Kubernetes is great to run stateless business-logic applications. But every stateless application depends on state, be it state stored in a database or state that is flowing over the network. Due to this complexity stateful workloads are usually deployed outside of Kubernetes. Without bringing both stateless and stateful workloads together under the fold of Kubernetes organizations cannot truly and fully benefit from the agility and operational simplicity benefits of Kubernetes. But due to the complex dependencies that stateful workloads have on the underlying infrastructure it is extremely difficult to plan for and run complex (storage)-intensive and network intensive applications on Kubernetes. Furthermore, once deployed, managing the critical day-2 operations on these applications is also difficult, including healing, scaling, updates, business continuity, disaster recovery, security, etc. Robin solves this through the 3 products. The first is Robin Cloud Native Storage (CNS). It is the highest performing application-aware cloud-native storage for any Kubernetes and in any cloud. The second is Robin Cloud Native Platform (CNP), which is a self-contained, open source Kubernetes platform optimized for storage and network-intensive applications. The third, Robin Multi-Cluster Automation Platform, is a web-scale platform for baremetal-to-service orchestration and automation.
Robin has the beauty to be both a storage and a data management product, give us the pitch about the product.
CNS is the highest performing (2x-4x faster than competition) cloud-native storage stack for any Kubernetes on-premises or in any cloud with special focus on application-aware data management and operations automation. We built our storage stack from the ground up instead of layering it on top of 15-year-old Ceph storage or 12-year old BTRFS storage engines. This gave us the ability to truly make storage application-aware. What I mean by that is that our storage stack natively understands the affinity, anti-affinity, data locality constraints posed by distributed databases and applications and automatically works alongside the Kubernetes scheduler to ensure the optimal allocation of compute and storage resources. We also have fine-grained control over the placement of data blocks on disks across servers, racks and data-centers. This allows us to do multi-volume QoS, an entire application-level and application-consistent snapshot, backup and cloning capabilities. Building from the ground up also allows us to have better control on IO hotspot detection and live rebalancing of data to deliver a stable and consistent performance in a long-running shared Kubernetes cluster.
Give us 5 key differentiators against competition,?
- High-performance (2x-4x faster than competition).
- Application-aware means your DevOps team who is managing the cluster does not have to become a Storage expert to manage complex stateful applications on Kubernetes.
- We do data management for your entire application, not just its storage volumes. Entire application means we cover an application’s configuration, metadata and storage volumes when taking a snapshot, backing it up or migrating it from cluster to another.
- We focus on simplifying the life of teams managing the cluster by providing the industry’s first application-to-spindle storage observability for Kubernetes.
- Brings a consistent storage and data management fabric to Kubernetes for any environment — on-premises or public cloud, for any storage (local disks, SAN LUNs, cloud drives)
CNS supports any Kubernetes distribution deployed on-premise or in the cloud as a transversal storage layer, is it something asked by your installed base and partners or a wish to create some differentiators?
We originally set out to build a strong foundational storage and data management layer. True to our vision of making infrastructure invisible we built the architecture to be truly agnostic of on-premises or cloud deployment. Our first customers either chose it for just on-premises or cloud. But in the past year as hybrid cloud has started to take off, our customers are using CNS for both the environments and also to move applications from one to the other.
You introduced recently a free edition, is it a response to accelerate adoption as Pure Storage and Veeam acquired some of your competitors – Portworx and Kasten respectively – and for sure it will have an impact on your business?
We have found that almost all of our successful trial deployments turn into production, licensed deals. The Free Forever version was developed as a way to eliminate friction in getting new users to set up a test deployment. It is still early, but we’re tracking a rise in downloads and engagements with our sales and deployment teams, so the early signals are very positive.
How do you see the storage segment of Kubernetes now? And Kubernetes role more globally?
We are seeing a massive jump in stateful workloads on Kubernetes. In the recent CNCF survey, about 55% of respondents said they are running stateful applications on Kubernetes. I am expecting that number to jump to over 80% in 2021. As a result, I think we will see a healthy demand for storage for Kubernetes and significant market growth in the coming years.
You’re listed on GCP marketplace but absent form AWS or Azure ones? Any reasons behind this or do you plan to be added them soon?
Our products have been qualified and available to our customers on AWS, Azure, GCP, and IBM cloud for over 2 years. Availability via the marketplace is just an additional channel through which our customers can download our software. We are prioritizing which marketplace to offer based on our customers’ feedback on how they would like to procure our software on those clouds. We have already achieved AWS Technology Partner status and are engaged with them on the certifications needed to elevate that partnership to Select.
What do you think of AWS Outposts strategy or Azure Stack? Is it a threat for you?
In recent years, there have been multiple approaches to building a hybrid cloud, we have seen products and services such as Amazon Outpost, Azure Stack try to tackle this problem. But going forward, I expect the hybrid cloud strategy will revolve around Kubernetes. The recent launch of EKS-D and Anthos bare metal underline the shift from public clouds to Kubernetes as their on-prem platform. We see this Kubernetes-driven hybrid cloud movement as a tremendous opportunity for us. All flavors of Kubernetes, on-prem and in the cloud, can benefit from the high performance storage and app-aware data management capabilities. Robin has the most robust and highest-performing cloud-native solution for hybrid cloud environments, including AWS Outposts and Azure Stack.
From a business perspective, could you share some figures?
We have been more than doubling our revenue in the last 3 fiscal years.
How do sell your product? Is it channel only? What about OEM?
We started with direct, however, in the last fiscal year, we have built strong joint solutions with partners to create reseller and OEM revenue channels. You will hear more about this in the weeks ahead.
What about the pricing model?
We have consumption-based pricing (pay as you go) on marketplaces with support for such pricing. For enterprise customers who are willing to commit to an annual subscription, we offer per node per year pricing model.
How many customers do you have? What is the split between geos? And between different uses cases? Kubernetes flavors? On-premise vs. cloud? And how many instances of CNS are running?
150+ customers with a revenue split between North America 70% and EMEA 30%, and 80% of our deployments are on-premises and 20% in the cloud.
What can we expect from Robin in 2021?
Our mission is to empower our customers to leverage cloud-native technologies to drive top-line growth and drive innovation. We have built the industry’s strongest performing and highly capable technology to bring data/storage-intensive and network-intensive applications to Kubernetes and cloud-native platforms. We are the industry’s first application-aware storage built from the ground-up to be cloud-native with petabyte-scale customer deployments. We are the industry’s first cloud-native platform vendor on which the world’s first truly cloud-native 5G network has been built. We have brought true innovations to the market in this space and have amassed an enviable 65+ patents in core areas of storage, networking, automation, orchestration, etc.
With this strong foundation in 2021 we intended to further add additional innovations in the area of hybrid and multi-cloud portability of complex stateful workloads on Kubernetes, AI-driven automation of day-2 operations and further enable additional low latency storage and network applications on Kubernetes.