The road to reinventing data storage from a totally fresh perspective
I’m a technologist. So for me, starting Hedvig was not foremost about building a company. It was about taking the advanced technologies that had been my world for the past 12 years and giving them new life in a realm of IT infrastructure desperately in need of reinvention. I wanted to make software-defined storage available to companies who otherwise would never have had the opportunity to benefit from them and do so in entirely new ways.
Prior to Hedvig, I’d had the privilege of working on two formative advancements in data infrastructure: I coauthored Dynamo at Amazon and developed Apache Cassandra at Facebook. While working on these, I kept asking myself, “What else could be done with systems like these? How can they be brought into the broader market?” Towards the end of my tenure at Facebook, I was aware that my experience with Dynamo and Cassandra could be harnessed to disrupt a very large market, like the storage market.
While I had the technical chops to drive such a disruption, we were new to company building. As exciting as it was solving hard problems to build Cassandra, starting a company has been an even bigger and more gratifying challenge. In three short years, Hedvig has grown from a concept to a reality. Here are the three key things I’ve learned so far:
- Learn first, build second.
The first hurdle: accumulating end-to-end storage market knowledge – quickly. My philosophy is that if you don’t understand the past you can’t navigate the future. In the same vein, if you want to disrupt a market, you must be respectful of the innovations your predecessors have brought to market. Storage made sense to me as an engineer, but a lot of core concepts and practical applications initially felt abstract. I wasn’t 100% sure what a storage administrator did on a daily basis.
So, I began learning about data storage and even used my vacations back home in India to complete storage admin classes. I also reached out to storage administrators and interviewed them. One of the first people I hired was working as a storage administrator when I contacted him via LinkedIn. We talked for four months about simulated tests and possible use cases while I was getting my feet wet in storage. He ended up joining our company in 2012 and continues to drive customer solutions and installation of Hedvig.
Stepping out of the engineer’s mindset and into the shoes of the storage admin was the first step in accumulating practical know-how. The next step in my explorations of Hedvig’s potential market was taking my first-hand experience in the world of the database administrator to the next level, so I spoke to a lot of DBAs. All in all, I spent three years in delta mode, researching, connecting the technological dots and developing a clear path of execution for Hedvig. By focusing on learning before building, I got to really understand the archaic things that were being done and could contemplate how they could be accomplished in a more sophisticated way.
- A startup founder’s wardrobe has an endless array of hats.
Adopting the myriad roles of a startup founder presented another learning opportunity. For years, I had been laser-focused solely on product development. By contrast, leading Hedvig my focus is prismatic, touching every aspect of the business. Wearing an array of hats means that I now spend each day context switching between R&D, marketing, sales, operations, finance, trips to Costco, light bulb changing and recruiting—lots of recruiting.
One of my favorite hats is the manager hat. I get to learn from incredibly smart, passionate, and unique people every day. Everyone approaches situations differently. Exercising patience and showing respect for these unique approaches has gone a long way in building a responsible and inspired team. I also believe in full transparency. Every Hedvig employee knows about every aspect of our company.
- When challenging the status quo, help people believe resistance is futile.
At first when we started floating the idea of Hedvig’s solution out to prospective customers we encountered resistance to change in the storage industry. This wasn’t really surprising, as storage has been operating in much the same way for the last 35 years. People warned us that enterprises would want to stick with the same solutions they’ve had for 10-15 years and that storage decision makers would be too conservative to purchase an entirely new architecture.
It turns out the fearmongering was unfounded. What I’ve learned from many subsequent conversations with CIOs and CTOs is that an overwhelming number of enterprises are actively looking to break away from the old solutions but have yet to be presented with a compelling alternative. They have modern applications and infrastructure architectures driving more and more data, but the old modes of storage are simply incapable of bringing this to life. Put another way: Storage is now a distributed systems problem. Exponential growth. Untenable economics. You need a distributed systems approach to truly solve customer’s data challenges.
We are at the pivotal moment when a transformation in storage is inevitable if enterprises are to progress. The hyperscale shift has already happened in the world of databases, and the confluence of megatrends —such as cloud, flash, and big data— necessitates that the storage industry transform or break down. Traditional storage is a bottleneck that keeps companies from innovating and bringing new capabilities to market. For companies to remain competitive and keep pace with market change they’ll need to adopt new storage architectures to work in lockstep with all of the other modern application architectures they’ve already adopted. The truth is, clinging to old storage models will spell demise for businesses in this decade. Conventional architectures cannot address contemporary use cases without impossibly high cost and businesses are realizing that the economics just don’t work. Resisting the evolution will hold companies back.
At Hedvig what we’ve created refits the traditional storage paradigm with modern thinking. The problem we set out to solve and the solution we’ve built is what the market requires be built. It embodies the approach enterprises are looking for to make their storage environments infinitely easier to manage and affordable to maintain. This is validated by the overwhelming interest we are experiencing from early customers. I am proud to say that we set out to disrupt the software-defined storage space and all indicators point toward success.
Let us know if you’re interested in learning more!