Momchil Michailov: To the earlier discussion around big data and analytics, it would be awesome to be able to run the on-premise big data applications and then move that data into the public cloud. Then, rent cheap analytics and cheap compute from the public cloud vendors to do the analytics tasks that are CPU-intensive. They are not as storage-intensive. There are some very compelling arguments around doing that and being able to create a hybrid. Our focus right now is running platform on premise. Our platform does run both on premise and at the public cloud. We can certainly span that. We are very much looking forward to getting the infrastructure in place to be able to deliver these types of architectures to customers.
Sramana Mitra: Very interesting. Momchil, I’m going to ask you one set of questions before we close, which is about your company. You said it took you 14 years to build the company. Tell me about the genesis of the company. Where did you found it? Where are you today in terms of the evolution of the company?
Momchil Michailov: Based on experience, it’s all about being at the right place at the right time. A few of the co-founders of Sanbolic ran a company earlier called Number One GM, which was focused on broadcast and media production companies. They built out a business around workflow and workgroup video editing and post production. That company basically grew through a transition of non-linear video editing to digital format. Moving into digital format requires a lot more storage. It requires capability to have teams work together and access that data at the same time. We sold the company in 1999 to Autodesk.
Being focused at the application layer, we realized that the storage layer lacks agility, flexibility, and the ability to be accessed concurrently from multiple workstations and servers. That is how we started Sanbolic. We started Sanbolic with the idea of data and storage management and providing shared access to data at the same time. This was the early days of fiber channel. There was a competitive technology called SSA from IBM. It was at the beginning of storage area networks and the distributed computing architecture. At the same time, we got involved with Motorola and Nortel and started looking at the next generation data center Baby Bells NOCs (network operating centers). If you fast forward to today, 14 years later, that really is the public cloud providers.
We began, originally, as a shared file system company. Because of our background and connections with Eastern Europe, we started the development team in Eastern Europe. That was a cost-effective way of going about it. At this point, Bulgaria is the Silicon Valley of Eastern Europe. VMware’s second largest development office is in Bulgaria. HP’s largest EMEA office is in Bulgaria. So when you land at Sofia Airport, you find these huge complexes that are very high-tech focused. There are also a number of really good universities that provide plenty of talent for development teams. We’ve been building this distributed architecture and it was a convergence of experience from our past careers as well as the migration of IT architecture – fiber channel and distributed architecture.
What’s happened since then is that the economics, the development of compute, and certainly flash over the past couple of years has really pushed the envelope. It brought this notion of distributed architecture and that whole hyperscale architecture into precisely what we were developing. It is being at the right time at the right place .The market is coming towards us in terms of everybody looking for that hyperscale capability rather than a proprietary web tool workload they wanted for general enterprise purposes. That’s what we’ve been able to do.
We’ve been able to do that very cost effectively – being able to do it at a fraction of IBM’s cost in trying to develop a similar product called Storage Tank. We were able to develop in Eastern Europe and have access to this significant talent pool. They almost invested a billion dollars in that program. The product never made it to market. We have north of 850 customers at this point. Some of the largest enterprise customers in the world are on our platform. In that regard, we were both lucky and at the right place at the right time.