David Yakobovitch
Welcome to our newest season of the HumAIn Podcast in 2021. HumAIn is your first look at the startups and industry titans that are leading and disrupting NL and AI, data science, developer tools, and technical education. I am your host David Yakobovitch, and this is HumAIn. If you like this episode, remember to subscribe and leave a review. Now onto our show.
David Yakobovitch
Welcome everyone back to the HumAIn Podcast. This week, we are pleased to talk all about data and the rapid scale of data through enterprise companies through startups and the entire industry. With us today is Ken Grohe¹, who is the President and Chief Revenue Officer of WEKA². Ken, thanks so much for joining us on the show.
Ken Grohe
David longtime listener. I can’t believe I’m actually on, I’m so honored to be out here and talk to all your listeners and hopefully they add value to your day in some small way.
David Yakobovitch
Thank you so much. And in 2021, everything is about data. We’ve seen as we’ve emerged from the pandemic, how every company is scaling, every company’s a technology and a data company. So I’d love you to start with our listeners. Tell us a little bit about WEKA and why now is a great time to get involved in the data space.
Ken Grohe
I’m glad your listeners first chapter, technically by nature, if you’re listening to your show, which is great, and because I listen to it as well, but WEKA as you probably know, and some of the folks that might be data scientists listening in, they had to strip a wekabite. So it’s 10 to the 30th power. That’s a good way to future-proof it. It’s all you can fit in a file system. And some of our friends who worked at Google probably know how the origin of Google came out from 10 to the 100 power, but WEKA does a great job as we’re actually a limitless data platform.
So a new way to do storage. It’s all software, it’s all service subscription through the people you’re buying from every day. So if you run it through AWS in the cloud or on premises with Hewlett Packard, it’s a great way to get things done and solve big problems. Specifically, the reason we’ve been in the news so often is let’s say you’re doing one of the biggest races of all time and you have to have, I don’t know, a vaccine for COVID in your hands.
And you’re one of the companies that are, they’re household names that are going through it and disseminating it the COVID vaccines in place.
If you were to do that, and you had tens of thousands of scientists working on a project, just like if you were having tens of people working on a document, you wouldn’t use Microsoft Word. We’re as a place and everyone has a computer, but if you’re going to collaborate, use Google Docs.
It’s just a way that people can have permissions, version control and overwrite on top of it. No different from that parallel file system. So, what WEKA is, is a modern and limitless parallel file system, that’s easy to deploy any scale in the cloud or on premises for the people in the data center, solve big problems, whether it be making the vaccine or figuring out who gets the vaccine first, that’s really important.
So we’re working with some of the CIOs and the federal government to do that right now, which is huge or some of the cool things in AI, I Hope we get a chance to talk to it. Like autonomous vehicles going down the road and number one I don’t think we’re offending anybody by saying some of those autonomous vehicles, David, like, Tesla get rated the highest by JD Power and Associates and are universally regarded as one of the highest, most safest vehicles.
And they have the ability to actually go up and down and actually have asynchronous transfers across those 220 sensors on a vehicle, up to a satellite and then navigate around a skunk in the road or a yield ahead or whatever it may be out there. And then some of the people, they might say, ‘I’m seeing AI use every day’.
I’m seeing an Amazon Go store a pop-up or I’ve been to an Amazon Go; where you actually go in, some people call them whole foods. Mine’s still a whole foods, but you buy some great produce, probably organic and you walk out and you pay a little bit too much for the fruit. At least we do. And then you’re on your way out.
You no longer waiting cash register, you walk right through. You flip off your iPhone. I choose an iPhone. It works with Android too. But you walk out as part of it. And there’s all the lines of financial services on stopping false positives and stopping fraud as part of it. But my point being is we happen to be in the news a lot because the COVID vaccine happening in the news a lot about COVID resiliency, but it’s all facets of AI.
And I actually think some companies in the fortune 2000 and the Russell 2000. Are you using this time? So people said with chaos comes opportunity. You probably heard that before David, they’re taking this time to double down on the competitive advantage and its AI initiatives across the whole business, would it be retail, manufacturing, etcetera. And there’s some of them who are choosing are, most of them are choosing to use records, a competitive advantage. So it’s a great time to be here. Welcome.
David Yakobovitch
I love the name. I love talking about the scale of #data. At most companies you think about gigabytes and terabytes and even bigger into the petabytes and exabytes, but you’re taking a one notch up. And so thinking about data, it sounds that the amount of storage and compute, it just continues to scale. And perhaps the thesis is we have not reached peak data yet. It’s only just the beginning.
Ken Grohe
That’s true. It’s funny. You mentioned that a lot of people ask us your average transaction. What is it? you’d be amazed. It’s a petabyte and it’s not replacing your core general ledger or ERP information. It’s a brand new AI initiative. You’re taking this information and I’m sorry to reveal this, but Forrester shared some data and I see backed it up in my friend, Kathy Ireland, when she interviewed me, said the same thing.
She goes, is it true? Ken still at 71% of corporate data goes unused, genuine how much money was spent to create this information data. And it’s going on use. So that’s amazing. So the average sale for us is a petabyte and that’s two third of the time. It’s on premises. One third of the time, it’s in the cloud every time to go between the two.
So if your users are listening right now and you listen to listen through it is hybrid world real. Hybrid cloud is everything. In fact, many of our larger life sciences customers, I something for the data center, move it up the cloud, same date, and then restore it back to the cloud and then do a mixed workloads based on whether it’s better economically or throughput wise or IOPS wise or latency wise do on-premise or out there.
And then I won’t say what part of the US federal government said this, but when we told them our average sales, a petabyte and our largest one and production was over 170 petabytes, he said, that’s cute. Our sale with us is going to be well over an exabyte. So the scale is huge, but more importantly, what you can do with the information.
So, in the best test of time to bring it back to COVID is, putting the information and having all the scientists work through it and work through the genome structures work through the actual muscle tissues and the reverse as far as actual physics as part of it and the actual science as part of it, to actually figure out the way to get COVID for a treatment.
And then more importantly to actually disseminate the right way across the different jarred piece certainly makes an awful lot of sense. But, we live in a new world where you can actually have someone in your show talking about, I sold an exabyte and we deployed an exabyte. And our average site is, you know, between 50 and 70 petabytes or every sale, a petabyte, it’s a whole new world to go through it.
And I remember, my first job was working at EMC. I was one of the first inside sales reps in EMC, my daughter in 1988. And we were selling. I’ll never forget it. It was 256 megabytes. I’m sending it back to a one Meghan to make boards or memory arrays for assistant 36. And we were also sold 256 kilobyte and five minute 12 kilobyte memory boards as well. You look up much things have changed and basically done career right, 32 years. But the ability to harness data, he’s going to make or break companies and make or break individual careers. In my opinion, David.
David Yakobovitch
And now Kim, prior to the pandemic, a lot of the companies were all in on just on-premise or all in only on-cloud and multi-cloud strategy. But we’re now hearing this emergence, as you mentioned about hybrid cloud joining the two together, why do you think that’s so relevant today?
Ken Grohe
I can certainly think if you’re in a university and at the end of the day, you want an AI project and I’ll cut to the chase, not just for the greater good, but to recruit great talent. If you’re Harvard or Yale or Princeton or Stanford or MIT or Oxford, or these incredibly universities you need, the phrase publisher perish, to come out with great AI projects that use, you never endeavor comes into some type of cryo-EM microscopes and sums up the huge work that only large really universities would spend their time on.
So when you’re doing that and you’re recruiting that type of talent, you’re putting it into action. And that’s probably going to be on premises. So people can see it. People can look at it and then you have for the fifth year of life for the highest latency, highest throughput, you probably sit on premises, but at some point, people say, wait a minute, my resolution for vaccine or my resolution for COVID resiliency or my resolution for the hot Dana, we’ll call it the first petabyte that’s going to be on premise.
But a lot of people saying, wait a minute when things get warm. When I believe in a recency bias. So whatever’s been used the most. That’s where you’re making money from or making progress from. But David, what I’ve found is, and our customers have shown us this when the data moves up to a warmer level and you want to save some money, you still want it.
No one deletes data is data’s power. It’s the new gold is the new oil is a new source code, but when that data goes into less hot or warm, There’s some economical ways. And we partner with lots of companies. You do as well, that IBM does a great run. Quantum does a great one scality out in etcetera, that has some on premises object store, and you get some cheap and deep storage that you have the warm data available.
What all of our clients, and you said it better than me, a hybrid strategy. All of our clients have said, wait a minute, put that up in the cloud. So for backups restore the archive to go into the cloud on it. Cause you can wait a little bit of time to retrieve it, but you don’t want to lose it. So kind of a trifurcation of how data’s on premise.
Usually some of the object storage is very adjacent to it, very close, and then long-term backup recovery archives long-term use. And I put in the category for acquisition and target and maneuver is actually in the cloud. Now, sometimes people go all the cloud all the time, but I do believe the ability there’s not as much switching costs and moving the data and having a data mobility feature that goes from on-premise up to the bottom and back. So if you look back to your life, I have an iPhone X or an iPhone 10, some cutting edge people might have a 12 years ago used to get the phone for free, but you had to sign up for a three-year contract or geocoder excuse me.
Now you don’t know how to pay 700, 800 and $900 for the phone and still do a two-year contract. Or maybe some of the carriers don’t have it. What’s changed is the switching cost is easier. There’s less barrier to entry. So the switching costs between on premises to near on premises. We’ll call it near line to offline or up in the cloud. Actually it’s online, but off premises, it’s easy to switch. And with solutions like WEKA, not to call us a glorified data, mobility, plumbing. But that’s what we really are. We allow you to put the right data at the right place at the right time to get, manage your information across the entire life cycle. So you make the money when you need it, and then you don’t lose it when you really want to protect it for data protection. I hope that helps answer the question better.
David Yakobovitch
It does Ken. And when you’re talking about the scale of data, the petabytes and exabytes, when we’re thinking of both cloud and on premise and the hybrid combination, a lot of what we’re seeing today is the need to compute faster, the need to have a distributed nature.
And that’s a very exciting moment. For running systems in parallel, but not everyone has thought about those use cases there. Still at the beginning of, you think classically machine learning AI, distributed nature, but can all companies take advantage and get benefits out of distributed compute?
Ken Grohe
Great question. If you don’t mind, I’ll tell two real use case studies I’ve been on other video blogs. And they’ve told me that you remember the case studies really well. So a quick one, we supply to Untold Studios. They’re not Netflix, they’re not Disney, but they’re an up and comer. They’re a studio in the cloud in fact, Amazon web services promotes them quite well.
What they did is they helped get the Sony PlayStation out. That was a special product, very hot. And it just came off of the holiday season here in the United States, but that was very successful and they did some incredible immersion type. CG type, commercials, 32nd commercials in the States on ESPN, every commercial in the past football season, which is now in playoff seasons was all about it. So to answer your question, we supply to Untold Studios. Their competitive differentiator was they did everything through AWS and their talent base is geo positioned to everywhere. So they found a great artist in Portland, Oregon, sign them up, New Zealand, sign her up. Somebody else in Belgium didn’t matter.
So where you live in your hat in a COVID world, didn’t matter. They kept going. When you think about it, traditional Hollywood shut down during the beginning of COVID. Cause you couldn’t break the unions, you couldn’t get the talent, the labor, they, Brad Pitt’s Reese Witherspoon’s to go on site, you couldn’t see, you couldn’t create any of the content we watched. The tiger came and things like that. But what I’ve told students able to do is enable them to create content. Get the Sony PlayStation out. There were still some supply challenges there, but that was hardware based, not software based, but we love talking about Untold Studios. It is a great example of a cases are out there is another quick one smaller.
And I don’t want anybody listening to me get intimidated. Like, I was talking about the US federal government. He’s talking about Thomas drivers. He’s talking about people that are curing cancer or Alzheimer’s, it’s not for me. I disagree. Some of our clients literally can clip use agedness tomorrow.
One of our smallest on-premise counts is Oklahoma, a research medical facility, and they’re a completely charitable foundation. In fact, Michael Dell, quick shout out to you. You actually donated their servers and you donated their network, which is an incredible specific individual. They knew that for them to make research, although they love donations and the cost of the donation exceeded every donation.
We all know it’s free, David. But they said in this particular case, the need to have a parallel modern file system with no limits, no compromise. It was so important because you’re going to bring all these engineers and all these scientists, you want to make breakthrough discoveries. So we were honored to get the dedication from Michael Dell and the servers, and your honor, to get the donation of some of the network gear, It was Mellanox.
I believe through Nvidia now owned by Nvidia, but the parallel file system was so important for them. They actually invested in the bought WEKA. And they’re a great reference for us. And I appreciate it along with GGN and the city of hope they’ve talked about it as well. But my point being is that sites started with only 64 terabytes, 5 years ago, we’d laugh like who could ever fill up 64 terabytes.
Now I’m here as an executive at WEKA saying that’s one of our smallest accounts on premises on the cloud. It’s a lot smaller than that, but that account actually yielded some dividends. We just got a phone call, not long ago, they’ve already completed that. And they were actually at an object store very recently to that. So it could be as small as you want it to be. And for real money, that might be $50,000 to make that happen. So these are not all multi $10 million projects, some are, and those paid a hell of a lot of bills and that is more engineers, but some are quite small and very nimble. And I do think you’re able to have a competitive edge, whatever size company you are.
David Yakobovitch
Tying the story together throughout your career. It’s extremely fascinating to see how you started your early days at EMC with small data or at that time, big data. These kilobytes and megabytes of data has today into the 2020s. Become a new story of the petabytes and exabytes. And that’s also spawned the rise of a new title, a new role that we’re seeing, everywhere from startups to fortune 500 companies, the chief data officer, there’s a lot of benefit to it, right around security, around control, around strategy, whether some of the important features you’ve seen or the needs for the rise of this new role.
Ken Grohe
Thank you. I’m hoping there’s some young I’m trying to pay it forward and hopefully there’s. Some early in the career, 20 something, it says what am I going to do with the rest of my career? I heard AI is great. I’m telling you now the chief data officers where to learn. And as part of it, you may not earn that job right away, but think about, and put this individual’s going to be, and typically they’ve come from the HPC high performance compute environment or David, as you eloquently said, the academic environment.
And what happens in those environments cause they’re a little bit static right now and mostly they’re doing zones or other type environments. So they’re not really thriving given the children and the students and the faculty are not on campus. What’s happening is some are being, let’s say recruited by the Fortune 200, trying to press like a bit of advantage.
And if you look at it and you could look at the stock market through all capital markets throughout time coming out of a tough period, the bigger tend to get bigger in the short term. So what’s happened is a title has risen. It’s called chief data opposite. As you said before, some of it is compliance and there’s certainly a chief compliance officer in there, but more important, more exciting is building out new applications that grab market share and new revenue streams using that.
What I say before 70% of your data corporate data was unused. Going through those huge corporate jewels and find out what’s really important for them and put them in other uses and new customers or new places during the period of time with the chief data officer has got the commission to do is be the bridge and the liaison between what I would call the HPC or the data scientist, which to be fair are always thinking about what’s possible. They’re the ones who were when John F. Kennedy said, we get to the moon by the end of the decade, when he said that at Rice University in the 1960 timeframe there, the people said, I know how we can do that. No one else figured it, but they could dream and make that happen.
And ironically with Martin Luther King, I have a dream speech there, the people that dream and know how to make things happen, maybe on a different time frame than the corporate America. Now convincingly IT it’s called a CIO. They have the rigor and discipline that we’ve all been used to. When you’re a new employee, you love the fact that you provision a new computer and it works password protection and encryption.
And no one’s going to break through that network. The CDO is good, the unique skill set of bringing the best of data scientists, the best of corporate it. So some people can say cool, but it is safe, but slow. Some people say, I love what these data scientists do, but it’s too abstract and you have to put some time parameters on it and we have to put dollars and cents behind it and ROI behind it. Boom, the CDO does both.
So David we’re blessed that that’s our persona and that’s the change agent that we’ve been selling to. There are people out there saying I have a storage refresh coming out there, or some storage coming out of a lease. What could be a great solution? What we’d rather be is the Ray of Greenfield, a brand new environment where you want to make sure your money with and put your chief data officer in charge of that.
And literally I’ve spoken at a few conferences, namely ThoughtSpot, and a few others where the actual chief data officer type commissions and trade shows and larger events. And that’s the type of change agent that is not just investing in something like a modern and limitless parallel file system in the case of WEKA. But they’re beginning to change and the people get promoted within the companies. That is out there, usually from academia or from the HPC has the technical rigor that they have the credibility for normal IT, so the CIO is trust them, but so do the data scientists and more importantly, they’re the people driving the new revenues from the CTO.
David Yakobovitch
What’s so unique about everything that you just shared. Ken, it’s bringing the theme together around storage and the theme around data storage is everything. And if you can’t successfully store your data and have the speed, then your applications aren’t resilient. They’re not running fast. They’re not running always online. They’re not running whenever and anywhere. And so, the unifying theme I’m hearing today is that storage is so important to how our modern AI initiatives. And modern applications. Can you share with us a little deeper dive about your take on storage and why that’s important today for applications?
Ken Grohe
Someone once said, if you want to stop a car, the car doesn’t stop, the tires on the wheel stop the car, see a bit about good tires and good advertising by Goodyear. What do you think is closest to your data, but the storage? So I realized with all due respect, it may not sell as many magazines or as be on the front cover as many magazines as it did years ago.
Sorry. I was, you mentioned before I was humbling one of the first employees, employee number one 10 at EMC. When I first started in the business, you bought a computer. Imagine this world you bought a computer from, or a super mini for me, the IBM deck prying, Hewlett Packard, or Wang. And they all had the better attributes.
Lang was great and word processing, for example, and scientific people tend to buy Hewlett-Packard. If you’re a business person, IBM, it didn’t really matter. But you bought one of those. Unfortunately, how EMC came about was it was unfortunate when you bought from that computer. It kept it captive on what memory or storage you bought from.
So imagine this and everybody who bought a computer knows you get to upgrade every three or four years, or because just the ability around you, everything changes. So imagine buying this unhappily company and being told you bought an HPE computer, you have to buy HPE storage, or you have to buy HP memory. That was suffocating. You were chained to it. And, 90% margin on the product. And they weren’t trying to be very innovative. So the birth of a company called EMC came from that, they kind of define the storage business. I don’t know who’s on the Mount Rushmore of storage. That might be a whole show for your day to do in the future, but I guarantee you the Moshe and I, or Diggins two of those four, so did basically create an EMC.
And I was blessed to be employee number one 10 as part of that. And we made storage, not an afterthought, but a forethought, because you can think about it just like when you brake on a car, the tires. And the friction of the tires stops you and saves your life, not the car. The intelligence tell the tires of the reels we show the tires of the braking system as part of it, but it’s really more important than you think about it. Just like a fishing line, as far as the actual fish.
And when you’re touching that, if you’re so inclined to do that today, but as far as storage for us, EMC kind of made storage relevant. And I talked to my brother, he works at Dell and he tells me a big part of their quota. And he has the whole quarter for a big part of the Midwest. A lot of those still store even more important than the clients or the servers that drive, which is impressive. But I think work is doing such a Renaissance with storage that the dedication to it and the proximity of the data is so critical.
You’re going to see a move of moving that storage from an external array one dated you’ll see some represent architectures where you actually moved that data. That’s on the storage closer to the processor. My friends at SAP have already done that. If you buy SAP HANA or you read it in the magazines, all they talk about is SAP HANA. It’s any word, memory processing for SAP. It’s a way you get it faster and they let the long-term processing go to the cloud. Or an object store makes sense. When you’ll probably see our partners at Nvidia, you’ll see the ability where the storage actually comes, becomes closer.
We haven’t talked about GPUs that much, but graphical processing units, I’m bringing hammers anymore. In fact, most AI initiatives, that are, whether it be Nvidia or AMD or some of the other GPU environments that are out there, you’re going to see the proximity of storage getting so close and actually might be in the GPU long-term.
And that’s a killer app that would make storage even more relevant. Given the proximity to the server and processing. It’s not just the size, it’s not the cheap and deep it’s how fast you get things done. So there’s a book by Bill Gates. We’ve already put speed at visit the speed of thought. We’re getting pretty close to that. And then some of our best clients, especially those in New York City will listen to me on the hedge fund business. Nanoseconds made differences. Have you ever read the book flash boys? There were reasons why they actually put carriers in conduit from Chicago all the way across, Hudson river in New York city to gain nanoseconds and microseconds and nanoseconds as competitive advantage. But, storage is going to have a Renaissance or is we’re living in right now, part of AI.
David Yakobovitch
I like how you bring up the different companies that are involved, not just in storage, but also in compute mentioning about the Nvidia. And we’re seeing that it’s bringing both of these together. In fact, the convergence of storage and compute of taking things like flash of GPU and cloud. Why are we positioned as a society in such a unique time for this convergence?
Ken Grohe
There’s really three metrics. And then you sound like one of our sales presentation. So if I was mainly the CIO recently, it’s almost like you’re a bug on the wall. My God, you’ve done some great research on this, but really three things have really changed. At the end of the day GPU’s are more prevalent than ever before. We actually can talk about you do a whole session on what GPUs can actually do. And I mentioned before, one of the best use cases we’re in the case of Amazon Go, just turning out whole foods, real place that doesn’t have a friend I’ve registered system.
That’s a reality. That’s something I’ve experienced several times and it’s really kind of cool. So you’re seeing the preponderance of GPU’s B everyday life and creating a new reality for everybody on the compute side. On top of that, you bail out the plasticity of cloud is huge. So you don’t need to be a giant company with 10,000 employees to leverage the cloud.
You can date and you and I can start a company. This is great. About America and about an opportunity we could do it. And you get started a little bit is as cloud as possible. And the last thing is NBME the ability to actually harness the actual flash is huge. I used to run the flash group for ANSI before I left. I was part of the general manager as part of doing that. But the ability to harness that is huge. I’ll give an example. I didn’t know this, but one of our friends that I didn’t see told me this 92% of all the Nvidia devices sold actually go inside the server. So back to that thing from five minutes ago, everything’s getting closer to the compute proximity compute gets more things done faster, and theaters are their only computer.
You don’t have to go out to the stage to anything because maybe the stage you have so mild latency. So to answer your question, I see three different paradigms. GPU’s being prevalent. NBME being everywhere in the network, but especially in the GPU and the server itself. That’s powerful. Now of course, Ryan, the economics LS city of cloud.
What’s cool about records? We’re writing all three of those. In fact, one of our investors actually isn’t video mentioned that before. We’re very close with them. We have, we’re blessed to have some reference architecture for them across their 8,100 and other things to come out in the future.
David Yakobovitch
When we put this all together and we tie them into applications and building for scale today, storage and compute. We’re thinking about bottomless, right? Can we have an infinite amount of storage? Can we be across multi-cloud or hybrid cloud? And can a tie for the compute to get that distributed nature, to scale the applications on demand. These are some of the topics can, we’ve been discussing throughout the show so far. What do you think puts all that together? Why is WEKA very well positioned as a modern data platform today?
Ken Grohe
You mentioned the word bottomless. I liked the word limitless. I don’t want to have my manager of Iran. One of the co-founders of WEKA. Really likes, saying the talk track. I don’t want to have any compromises. I don’t want across the skill of simplicity or speed or scale. I don’t want any limits. I want in theory storage to be everywhere where it should be storage everywhere. That’s kind of our mission. We want to get to eventually, but to put it all together, what we begin to see, and I unfortunately happened to read an article fortunate for me, but unfortunate data point.
That 1.2 years ago before WEKA started coming up on a scene and Vinnie even got more prevalent and NBME got more prevalent. All the intended AI practices and initiatives, it was going to be a fallout that over 50% of them were not going to have ROI. And that’s unfortunate. Now that number has shrunk to less than 12% per the analysts we spoke with yesterday.
And I speak with IDC, ESG, Gartner almost every day. The reason why it went down and the lack of the success ratio got so high is people are believing. I’m going to go back to geometry class. Imagine the role of an equilateral triangle. All the sides are at 60 degrees. I know triangles. There are degrees between all of them, all the same sides.
Imagine if they’re all the way doing that what’s going to happen with that is you’re going to put a GPO in the environment or clustered GPS. And then you’re going to say, wait a minute, I can’t just do it all on compute. At some point, I have to burst it out, back to the clients, back to the people I should do in the work, the sciences, etcetera.
You’ve got to upgrade your networks. That’s why Nvidia bought Mellanox. It makes an awful lot of sense. And then on top of that, you’d say, wait a minute, I want a parallel file system too, because I don’t want to be waiting for the worst word for that word, but it should be George Carlin’s bullet a word latency. You never want to have strengthened aptitude and intelligence, but you don’t have the ability to use it at that time. So the pro file system lets everyone use it all the time. We take care of the locking and the overriding, all the other management is part of it.
So everyone can flourish all the way through, kind of like a Google doc versus using Microsoft word. So going across those three corners, if you have a balance between the GPU fast network Mellanox, or maybe a Rista or a tour gigabyte pouring here, my tech network or InfiniBand academic world, and then you have the parallel file system. You may have all three work in concert.
That’s where it happens when we’re hopefully saying to everybody, David is, you can start as small as you want and go as large as you want, but bring the ability and the imagination to solve big problems. Because storage and more importantly, AI centric accelerated storage from WEKA is certainly huge. And I love I’m going to use an ops shoot of your bottomless. I’m going to call it limitless. So it’s kind of the solutions of limitless.
David Yakobovitch
And, everything I’m hearing today is that with storage and compute, it’s not only on the cloud, but it’s also on device. You’ve mentioned about how our phones have changed over time. Today the average phone has gigabytes of data. I believe some of the new phones even come out now with one terabyte of data. So it’s fantastic. How we’re seeing that scale. We’re seeing #machinelearning chips on the Apple phones and the Samsung phones, which is going to enable storage and compute closer to the edge as well.
And so for a lot of companies today, it’s to know that you don’t have to be only on cloud, you’re going to be running on multi-device and that leads to a lot of data mobility. Why is data mobility, something that companies should be thinking about today when they’re building applications and scaling for their audiences?
Ken Grohe
One of the biggest speeches of WEKA and I came up, I didn’t come up with a new word, but our friends at Gartner gave me a new one. Last week is extensibility. I have to say in a couple of times, what that allows you to do is imagine the power that you just mentioned, data mobility. And I respect that, but you want the right data at the right place at the right time. No in all the cases. So you can capitalize, you can make, go faster and go actually press your advantage. And wherever it might be, whether it be retail or manufacturing.
The reason I say extensibility is for naming conventions, whatever file you create, you want that same name and convention whether you’re on premises, we on a cloud, we were an object store or whatever. And what’s great about #WEKA. You haven’t talked about it yet. It’s one global file namespace. So you’ll never leave the naming conventions and you never leave the actual policies as part of it. So. The company has a certain amount of rigor and we want to protect everything.
So it’s full five, 12 kilobytes protected and data at rest and in transit. And that’s what we do in WEKA. So we never put your day-to-day needs have a better ability, but it’s not just data mobility. It’s data of ability for the right data, the right place, the right time. So it’s kind of like an information life cycle management to have that as part of it.
And that certainly helps, but the extensibility is a nice phrase that Gartner gave us basically to what’s cool about WEKA. You’re in the same file system all the time. And what does that help you? At some point, coming out of COVID, there’ll be a company that’s big that buys a company that small or more likely David someone that’s fast by someone that’s slow.
When that happens, you do some due diligence, but eventually you get to merge databases. How cool is it to have one file system globally for all of that. And now you’re not saying, who hasn’t said that we’ll get the report to the next year. We’ve got to sync our databases or working some type of batch work to go through it. No. Your time is speed. You can’t fight physics and you can’t teach speed. You’re going to aggregate that together.
And the acquisitions have to be done in days of not hours, not years and years or months and months. So, like I said before. The fast, eat the slow. If that’s the case, the ability to move the correct data, the right naming conventions based on the right policies, the right security allows you to happen.
So we’re kind of known as the information life cycle management type company. It does the global namespace and the kind of new modern storage way to do it. But thanks for asking a stupid question. Actually gave me a little bit of a jab there by saying, is it just over the mobility? No, it’s extensibility and data mobility, various two questions.
David Yakobovitch
That’s right. And I love the comment that you shared about the fast eating the slow. We’ve definitely seen 2020 and 2021 accelerate into a fast world, with the world this year rolling out vaccines with our recent inauguration of president Biden, we’ve seen a new opportunity right where technology is integrating with society everywhere.
And we’re moving into a space where the world can become a better place, especially as these vaccines rollout. But can you set the stage for us that what led to getting the world into the new normal as we’re going to be moving into reopening later in 2021? What some of the research that WEKA has helped empower for today’s vaccine and #COVID19 search rollout.
Ken Grohe
Cool. That’s very timely. And I didn’t get to see a lot of the inauguration, but, as an American, that’s always great seeing the transition and a very patriotic day to go through that. So thanks for acknowledging that. And obviously I’m okay. Earlier this week, as far as the day off on Monday, but to answer your question specifically, we were involved with the vaccine makers and the FDA submission.
That was very important for us. So, and one of the biggest races, at least in my life, I’m 53 years old. Depending on where you are and the whole parameter, as far as being involved by COVID in my family was affected by COVID my own brother had it. For example, we recovered pretty quickly, but he’s one of the lucky ones. Some people weren’t as lucky.
Not a flashboard. We were involved with vaccine development, obviously with those, most of those vendors did all suppliers do that through the cloud and we have a solution for them in the cloud but ultimately a hybrid solution as well. So as they move out of the data center and go to the cloud with records, therefore at St. Global namespace, but a part I’m going to talk through is because people always say, he must’ve given so large databases in such large files systems over to the vaccine makers.
There were large, but around 10 petabytes, actually, you and I talked about this offline before David there’s a company out there called Genomics England. What a great tee up. Thank you. They were racing to do 5 million genomes or the year 2023. And David, Arlene, his team were trying to do that and they were on that path. And then all of a sudden March happened. Everyone’s gonna remember March, 2020, where they were that period of time because everything changed.
It basically pivoted and said, wait a minute, we have one of the largest data lakes in the world. It was 70 petabytes. Let’s use this for COVID resiliency. I don’t want to get too much in the politics, but you can certainly understand the data. You’re certainly the same one as well because you travel internationally all the time. Certain parts of Asia were a little bit distrustful of where the cure is going to come from, or the vaccine is going to come from a certain part of America. It was certainly in a little bit of a distrustful environment. During that, especially with our last president kind of fueling those fires. My point being is we will work with Genomics England, and our partners, the Swiss and Swiss Institute of bioinformatics.
They actually housed the actual resiliency for COVID. Because it’s a very neutral site. Everyone says, Switzerland’s a great place to put that for banking and for staying out of warring factions. That was certainly great. So we have a great case study and the Swiss Institute of Bioinformatics, it’s a bit more our website, if you’re still inclined, it’s on their website as well.
But Genomics England was where they actually, the COVID resiliency. I happened to work with Tony Scott. The former CIO of the federal government was advising the Biden administration as well as others.
And there, they actually have a full plan that you’re seeing real. Tier one through tier six tier ones, obviously first responders, healthcare workers, etcetera. Tier two is obviously dentists and folks like that. And then people with asthma get into tier three, they actually did for the benefit of humanity, what would it be the best quality of life? And I just talked about this before they were 28 petabytes of my former company, Dell technology, and they were very happy with that #technology.
But if you do 5 million genomes divided by 20 petabytes, it doesn’t work it as they say before. Many doesn’t play that game. So they had upgraded 70 petabytes and here’s the good news. And you talked about Oliva before that company, Genomics England, or that organization, which is funded by very generous individuals.
And think about what they’re doing. They’re figuring out who gets a COVID vaccine at the right amount globally, which is great. They happen to have 10 storage administrators when they had the Dell technology, which is 28 petabytes. They were able to re-position those individuals to other people during COVID times.
And that 70 petabytes was managed by just one individual. So that’s counterintuitive, but it’s real 28 petabytes, 10 storage administrators to 70 petabytes with only one you might say to yourself, it’s great. That storage is growing and growing, but how are we going to manage it? When you have 70 petabytes more one individual, that’s the answer and be specific.
We also sourced to the federal government. They have been great to work with, but for all the right reasons during COVID, they had to work one week on one week off, and you want to talk about security. These folks don’t get the advantage of using the browser. They have to use a walk down terminal at the Pentagon with full clearance.
If you’re going to do that, and you’re working one week on one week off, you’d better make hay when you’re at work. And so the demands of running a large environment based on 2003 technology didn’t make sense. So whether it be GPS or a GPFS or luster spectrum scale, etcetera, it just didn’t come from them being upgradable rack up modern file systems. Because you were able to manage more with less time and more per person. So that’s kind of the answer. You’ve got to have a system that’s a global, single global namespace, but auto tunes itself. So you can’t be spending time doing Ben Files and London Management and things of the past. You’ve got to have it running thrive.
And he focused on getting new sources of revenue, new use cases and new competitive advantage and less on things that are taken for granted. And to be brutally honest, AWS and the cloud vendors have done that first off people assume to have uptime. You assumed of access work on the next way. That’s what record allows you to do in the next wave.
David Yakobovitch
So taking everything that we’ve been hearing today about the rise of storage and #compute the chief data officer, the movement into these NMVS and this exciting new compute technology and for good with COVID research, there’s been a lot of fantastic topics we’ve covered about the work at WEKA. Ken, what are some of our next steps or actions you’d like to share with our listeners today to continue their journey with big data?
Ken Grohe
We asked you to join us over weka.io. that’s our website. There’s a lot of information, a lot of white papers, a lot of solution bleeding briefs that you can read about, feel comfortable about how we have an ROI tool. We actually have that we just reviewed on there as well. A lot of logos, Microsoft, for example, a lot of huge customers.
And John was driving that you see, thank you for scrolling through that National Institute of Health. Obviously we worked hand in hand with them with through COVID as well, but feel comfortable, go through that or send me an email at grohe@weka.io or follow me on either Twitter or LinkedIn.
I’m a lot more active on LinkedIn these days than I am on Twitter, but please do it that way. And then we’d love to get you a free copy for you to try on AWS, if you’re so inclined to use it in the cloud, or if you will, for some reason, and you want to do it on premises. We do have some amount of kids.
We’re not limitless on the amount of kids we consider people, but we do have some of our great friends at HPE and Supermicro and Dell and attachment tower have loaned us some equipment that you can navigate, how they not a proof of concept, but a proof of value, POV kits, proof of value. So, if you want to stick your neck out and show how easy it is to gain market share and getting revenue with your company, we’ll send you a POV kit and you can actually make that happen.
They’re typically around 64 terabytes and bring us your biggest problems. We can make them happen. So whether it’s on premise or use our bowed partner in the case of AWS marketplace, we’d like to do it at no cost to you knowing convenience. You’d be amazed at how well it can all make it happen. We don’t have a WEKA button just yet, but I do use an easy button.
People have told me that they do make it pretty easy. And one of our customers had said, you should change your name, not to WEKA, but will play off. You can get it done in about a week. Whereas before you just take several quarters, if not years. So thank you for the nice tee up. I’d love to see you out either on LinkedIn or on Twitter or email. I threw that out there as well.
And obviously go to weka.io, there’s enough space out there that you can feel comfortable at your own pace. You can learn at your own level and when you’re ready, we’ll be ready for you. And, you need to know this, all the data. You’re a very technical person. We actually have four technical solution engineers for every one sales person.
So we’re not a very sales dominant culture. We’re all about solving big problems and very technical by nature to get into your use cases. In fact, most of our people spend most of the time trying to move data scientists or people represent data scientists. So if you’re in that category, we’d love to help you out.
David Yakobovitch
Ken, it’s been a pleasure for all our listeners, Ken Grohe, the President and Chief Revenue Officer of WEKA. Thanks for joining us today on the HumAIn Podcast.
Ken Grohe
Thank you.
David Yakobovitch
Thank you for listening to this episode of the HumAIn Podcast. Did the episode measure up to your thoughts and ML and AI, data science, developer tools and technical education. Share your thoughts with me at HumAInPodcast.com/contact. Remember to share this episode with a friend, subscribe and leave a review. And listen for more episodes of HumAIn.
Works Cited
Companies Cited