By Sanjiv Goyal Featuring Mohit Aron
In this article you will learn about:
- Trends in data management and storage
- Entrepreneurial opportunities in data
- Data lakes
- Ransomware
- Contemporary Business Model Innovations
- Tips for entrepreneurs
- Snowflake
Cyber Oil
AI is all the rage right now. Everyone wants a piece of it, but Mohit Aron remains focused on his first love, data. The IIT Delhi alumnus, co-founder and CTO of Nutanix, and founder and CEO of Cohesity, illuminates how data will shape infrastructure and industry in the future. In his words: data is the new oil.
In the last 25 years, our world has seen massive technological advances and been shaped by the ways the internet has connected us. As network speeds have become faster, as disc capacities have grown, as the cloud has evolved, data has exploded. Zettabytes of data are processed every year.
These days an organization can no more be divorced from its data than a smartphone can be separated from a millennial’s hand. In order to be on the cutting edge of business, you have to be in the business of data.
Next Steps
With so much data in the world today, storing it in silos that are connected via the internet has become costly and cumbersome. The next generation of data storage is looking to consolidate these silos and streamline the process of data transfer.
In the past, it was easy to move data because there wasn’t that much of it. Now the volume of data is growing faster than network speeds can handle. That means the future of data will be built around trying to keep as much data in one place as possible.
Hyperscalers like Google and Facebook have taken care of some of this consolidation for consumers, while companies like Cohesity are looking to do it for the private sector.
The developments in data right now boil down to two areas:
- Consolidating silos
- Managing data on platforms
Once the silos are consolidated, rather than letting data sit in them, it will be placed on a platform where value can be derived from it in new ways. While this is liberating and efficient, it also comes with a unique set of management and privacy challenges.
Ransomware
These days instead of kidnapping a person and asking for a ransom, you kidnap their data. Ransomware is used to break into a company’s network and hold its data hostage, effectively shutting down operations until the ransom is paid. Some very large organizations have fallen victim to this cybercrime including UCSF and Honda. Aware that data is the lifeblood of industry, hackers go for the jugular.
The Implications of Data Accessibility
A vaccine used to take decades to get to market. Just think of how long it took to formulate a Polio vaccine. We have significantly shortened that time frame because of data collection and accessibility. In the case of COVID, the Chinese encoded the virus’ genome early on and made that data accessible so labs around the world could start working on the vaccine. In as little as ten months the world had a number of vaccines being formulated. This was possible because of the internet, the processing power we have, and the vast amounts of data that we can store. It’s fascinating what we can do with data today.
Imagine if we had a platform for hospitals around the world to share data and the ability to ensure privacy and security. This may be closer to fruition than we think. All of the big cloud companies are building data lakes where people can house data no matter what part of the world they are in. Cohesity is working on something similar for enterprise data.
Reimagining the Architecture of Data Storage
The trend in data storage is to consolidate and reorganize, to take it from a hierarchical model to a flat architecture.
According to Aron, “People have realized that the world of tomorrow isn’t a siloed world, but one in which data is accessible to all. But you need to make it accessible in the right way so it’s protected, can’t be stolen, etc. The ability to access and run computations on that is a privilege.”
Innovations in this area will allow us to store and expose data in a safe way, as well as allow third-party services to compute it and applications to run on it.
In the future, instead of copying data, we will copy computations. In the future data sets will now be analyzed and modularized by third-party software. The goal is to be able to “wrap” a computation in a container and then send it safely from one place to another. Data storage companies will offer this as a service.
Aron predicts: “We can expect to see this trend picking up in the next five years as already giant companies like big banks are showing interest in creating a modularized way of running computations on their data, as opposed to the legacy way where all the data was in silos and we had monolithic servers in charge of running computations.”
Business Model Innovations That are Here to Stay
New business models have emerged from developments in tech and data. Consumers are being given more options than ever before. Take the example of Uber. We have the option of buying a car or using a car service. The Uber model exemplifies: pay for what is used. The consumer pays for the time they are using the car and nothing more.
The world of tomorrow is a hybrid world of bespoke multiplicity. “We are emerging from a world where we only had one choice, and we are going towards a world where we now have multiple choices,” notes Aron. With personalization and diversity sweeping all areas of culture, successful companies of tomorrow are going to resemble ecosystems creating win-wins for all through profit sharing.
Shifts in the enterprise sector related to data:
- Moving away from perpetual license models and making the shift to subscription models.
- Moving away from ownership toward third parties having the responsibility of ownership costs while providing a service
- Moving away from buying outright and toward paying only for what is used
Layering Is the Way Of the Future
The implications of layering become clear when we understand what it could mean in a real-world scenario.
I asked Aron if he thinks the major data collectors (government, private enterprise, and institutions) will one day cooperate and find a way to safely share data for the greater good of all. He maintains that in the future we won’t need to rely on their cooperation at all because building layers on top of datasets will make the need to cooperate a nonissue.
“At some point, a layer will be built that encompasses that data, and it will be about how that layer is regulated. So if you interact with that layer, it won’t matter if the three stakeholders and data holders in the layer below do or don’t work with each other.”
Hyperscalers like Google have taken the original silo layer of data and built a platform layer on top of it. When you do a Google search you don’t know where the result is coming from, and you don’t need to know. The third layer that will be built on top of that is one of abstracted insights that aren’t concerned with the layer below it. Being on a layer provides the user with a contract and hides the layer beneath it. This is what Cohesity is working on.
For now, data is mostly being stored and computed in the lowermost layer but that will change as more layers are built and APIs become the middlemen.
“We need a level of abstraction like the Googles of the world have built, not on an enterprise scale but on a mass scale. This layer of abstraction will sit on top. Now that you have an abstraction it removes the issues with the silos, and you can now provide a marketplace on top to work on that platform.”
The tech company Snowflake built an abstraction for data warehousing that sits above the cloud. If you interact with the Snowflake API it gives you a data warehousing abstraction. You are unaware of where that data is coming from but you can still utilize it. Additionally, you can build further layers on top of the warehouse layer.
Aron explains the concept of layering in simple terms: “There’s one layer that handles all the underlying storage. There’s another layer that makes it all look like one platform. There’s another layer that does something interesting on top. There’s another layer that implements email on top. It is a layered architecture.”
For Entrepreneurs
For the future of tech, there are innovations to be explored in every aspect of software. These are the areas that will call for innovation going forward. Security generally, but specifically associated with shipping computations and transmitting information.
Entrepreneurial opportunities in data:
- Security
- Privacy
- Government compliance
- Modular Data Transfer
Mohit’s advice to entrepreneurs today:
- Think like a futurist:
“If you’re an entrepreneur, don’t work on the problems you are seeing today, work on the problems we will see two years from now.”
- Be a team player:
“You have to move into a team mindset when you start a company. The sooner you make the transition from I to we, the easier it will be for you to become a great entrepreneur.”
- Be prepared:
“Don’t rush into starting a company, take your time, and hire great people.”
