Extension two twenty four. Give her a call. That number again is nine oh nine eight A nine eight three seven seven extension two twenty four.
You're listening to the Inland Talk Express ten fifty AM and one oh six point five FM CACAA, Loma, Linda.
The information economy has a rid. The world is teeming with innovation as new business models reinvent every industry industry. Inside Analysis is your source of information and insight about how to make the most of this exciting new era. Learn more at inside analysis dot Comsideanalysis dot com. And now here's your host, Eric Kavanaugh.
All Right, ladies and gentlemen, Hello and welcome back once again to the only coast to coast radio show that's all about the information economy. It's time for Inside Analysis. You're truly Eric Kavanaugh here, and I'm so excited to be talking about engineering these days, engineering for success. And what's one of the hot topics that we have, especially with all these electric vehicles and hybrid vehicles these days, are the batteries. And batteries of course are expensive, they're heavy.
We want them to run as long as humanly possible. No one likes charging their batteries up people like using their batteries, and certainly in the electric vehicle world, they are crucially important. I mean, with a hybrid, you got gas. With an electric vehicle, all you have is the electricity and that's in the battery. It's very heavy, it's very expensive. So what can we do to improve the performance of
these batteries. Well, guess what we've got. Jovanni Rossi calling it all the way from Milan, Italy, my favorite country in the world. I love Italy, and he is with a company called Electra. They are leaders and applied AI for battery packs. I'm like, what explain this? So, Giovanni,
welcome to the show. Tell us a bit about what you folks are doing and how you're able to use a digital twin of batteries to improve the performance, and not just in the design phase, like meaning for the next set of batteries to come on the market, but for batteries that are already out there being used. Tell us what Electra is and how you're doing this stuff.
Thank you, Eric.
So, Electra is an AI clintech and B to B software company that is like focused to unlock the full potential of Barry technology. So basically as was mentioning. We
have two different type of products. The first one is like a digital tween applications, so we study the barries, we understand what is happening, you know, between the berry with having a virtual replica of the battery where you can thanks to our software capability and AI models, test and innovate with the berry, so you can insert your different parameters. You can test out new chemistries or new models what you want to do, what is necessary to make a new berries, and so you can get faster
to your results. Basically, you know, when you need to produce a new berries, it can take up to ten years and cost like up to one billion dollars, and so of course you need technology as ours to decrease of course your time and also your cost into making new berries.
On the other side, we also have a set of software for.
Like you know, managing, optimizing and controlling very capabilities and so improving again the performance of the berries and providing valuable insights for the different type of users and for the business managing these barries so to make the battery lasts longer and perform better and so maximizing of course the reternal investment for the battery assets.
Yeah, that's really interesting. So and let's talk first about the digital twin and that environment where you have built out in a computing environment. Obviously the battery itself, what goes into it, and then you're able to allow your engineers to play around with different settings. Is it getting into the material science? Like how much detail can you give us on what's done in that environment and how you're enabling more efficient battery designs.
Ye, can go into more than one hundred parameters, you know, into the very spects, so it can be either at a cell level and so like you know, into the small great stuff in that sense, into until like the very pack. So it depends of course of what we want to do. But we have different applications, different modules in the platform, and so you can go sell to pack to pack levels so that you can test and innovate whatever is like necessary for you. Of course you
can change parameters. There are already some pre built AI model that you know, we have already designed and tested with different customers and understanding of course the trends in the market. But then of course we can also create some additional customized AI models to do some very specific operation that a customer may be doing.
That's interesting. So with all these parameters, you're allowing the scientists basically to get in there and just play around with things and see about whether it's the kind of materials that are used, how much of them are used, all these different aspects of the actual design process. And then in this digital twin world you can kind of
play around. Can you talk a bit about the accuracy, like what, you know, what's the range of efficacy with these tests when you do them in the digital twin world and then you see what happens in the real world. Is that something that you're capturing and managing over time the data around that.
Yeah, I mean like you can play around with whatever it's like, you know, within the platform of course, and also adding some additional bodiles and piece when it is necessary.
And I think that the main benefit.
Here is like you know, when you need to test out some berries, you know, and you can have this digital replica and you can play around with it, you can decrease the testing time, so you know that that
seems like a real deal. And the game changer perspective that is, like you know, you don't need to have like a lot of you know, physical testing, but you can do everything virtually and this can go up to you know, ninety percent of the testing can be reduced, of course in the best case scenario, but we have seen that an average can be of like reduction in test is like sixty to seventy percent, so meaning that of course you're saving up a lot of time and
also a lot of money, and so the scientists can also focus on some more interesting stuff. Let's say this way than just replicate the test over and over again.
You know, I remember talking to a gentleman, a very very interesting guy. He is an electric car racer, and we were over in Belgium for an electric car race a couple of years ago, and he was talking about this this effect when you float you can actually slowly you can actually increase the energy in the battery a little bit. Do you know about this? Like when you break our cost in an EV, the electric motor access kind of a generator converting the energy back into electrical energy.
Is that right correct?
Yeah? And you can say up plum energy and refiel that of the marry.
Yeah, that's pretty cool. That's pretty cool stuff. And then talk about the battery management systems I'm getting spanked by the sunlight here in my head, but try not to worry about that. Talk about how you can also work with the management matteries I'm sorry, the battery management system to improve performance there. What does that look like?
So basically, let's take the an electric vehicle as an example. You know what you do right now, It's like you have different information that are coming from the berries, coming from the driving style, coming from the environments, and of course all of these is like impacting the performance of the battery.
So what we do is like we take all of.
These information together, so berries, environment, driving style, and we put everything together, and thanks to our AI you know, algorithm, we can provide an output to the users, to the drivers in the sense that is like how you can optimize, of course your range. You can so extend your range in this sense we have been able to estimate that. Of course, the reduction of sorry, the average error in like you know, estimating the range is like twenty percent
fifteen to twenty percents depend on vehicles. And with our model you can go down to less than one percent error in estimating range. So this means that of course you have much more precise information in that space, and we also you know, and with additional insights and suggest that we can provide, you can also extend the range
and the capabilities. We have actually been testing these very recently in a cross country you know trip coming starting from Boston to Vegas, because we were presenting our technology at CS twenty thirty five, and we did it with like a Tesla cybertruck. So our founder, a core founder and CEO of Britain Martini, and one Copio drove all the way down from Boston to Vegas to demonstrate the capability for all our algorithms.
That's Alloyd.
Yeah, it's a very long drive.
They looked like seven days, so that's really intense sad this way, and we have been able to you know, showcase and demonstrate that we can extend the range of twenty percent and reduce the charges of around thirty three percent.
That is pretty it's a pretty good number, we say.
And in all of these you know, also managing and understanding all the data coming from the various environmental driving styles, and of course we have also capability to predict failures in advance, so we can have like in a product that is like one hundred percent safe for of course the driver and and for them of course vehicle producer in that sense that is also very very interesting, and all of these can be scaled when you think not only you know individual drivers, but you have like fleets
of vehicles or flit of electric vehicles, you know, adding all of these information and understanding what is happening within the so of course, what is like the state of hels of your battery, what is like the degradation of your batteries, and so how much your btrier can stay can last? You know, for what we define as the remain a useful life of a battery, you can actually increase the value of your asset up to forty percent. So that is like a way for both the drivers,
the OEM or the battery producer. And then you know, the fleet operators to have many information, many sites and so have a much better capability to manage the operations and so of course get the value out of it.
Yeah, well for sure. So I'm guessing what you're doing here is in terms of being able to optimize the battery life is you're you're analyzing the driving behavior. So does this person accelerate very quickly the drive? It very high speeds on a regular basis, do they go up
and down often in terms of speed. These sorts of variations will have an impact on the battery life, and so what you can probably do is detect those algorithmically, understand what's really happening, and then give some recommendations to the user to say, hey, maybe if you were to not speed up and slow down so quickly, you would extend I'm just guessing here that you know, if you haven't even keel drive, you're going to get better use of the battery.
Right, yes, and your mattery can last longer because it's not allly about range and performance, but it's like you know, if you drive better that sense, so if you use it battery better, your barrier can last for longer. Of course, we have some algorithms like optimizing how much the bettery can last. But of course then in this case, of course also the users like responsible of that. So you know, the two things combined that is like what can make
you very lasts longer? And if you very lasts longer, you know, it means that you are keeping your cart longer in a way, and that's also saving money in a sense, but also results of some environmental impact that
is definitely important in this sense. And of course you can scale these not only to electric vehicles, but if you think also stationary storage, that is like another application, usually much more for B to B side, let's say this way of leg for businesses, but it's also becoming quite popular for homes, especially you know, mentioning you know the B two B side, or like businesses, when you have some solar farms or wind farms and you want to store the excess of energy because you know, renewable
energy is very cool, but you know the shine and the sun, sorry, it is not always shining and the wind is not always ploying, so you need to you know, to store the energy when when you have the excess
of energy and then release it into the grid. And so having technological capabilities and aid I driven capability to understand what is happening in these berries, to manage them better and to optimize them, and also to predict falls because you know, having all of these information and it is one of the main game change capability about what technology is that we can predict falls up to three
months in advance. So this means that we can understand what is actually happening, and you know how your berry is performing, and if you're going to have some critical faults, and when I mean criptical faults is like you know them a runaway or like fires or some other issue with a very that's very very important to maintain safety and reliability of that asset.
Yeah, no, that that's very interesting. And there are also these power walls, right, doesn't Tesla don't they have a power wall? So basically you can buy a car, but you also buy this wall that is like in your house, and it's it's it is a long term battery storage is what it boils down to.
Right.
It allows you to I'm guessing to charge the car faster and do things of this nature. Yeah, you also just have a power source, like you mentioned, because I looked into solar panels years ago when we were in Texas, and you know, some of the key things you have to keep in mind are being able to manage that power that comes in and storage such that at a certain point you can then pump it back into the grid, right, which is you know, you have to be very delicate
about that obviously, because it's electrical energy. You don't want to to get too much So what are you able to do there in terms of understanding the sort of dynamics of how battery usage works in these walls, and like, can you I presume you can extend their life too, right.
Correct, we can extend the life.
You can monitor them of course better and especially understanding how they are performing and increase their again performance and optimizing them so that you can have like more energy to use in that sense, and charging faster your car if it is a car for application for a car, or like you know, managing better d energy for your home. Optimizing of course also the a flow of energy because usually you have pav panels and so you want to have everything that is working in a very optimized way.
That's very interesting. So do you get much traction in that space, or maybe talk a bit about your clients who's using this right now? To what effect?
I mean?
Obviously it's you're just extending the life of this asset, and you're improving the use of this asset, which is incredibly valuable and saves money and also saves you on trouble, especially this part about being able to predict fires, because that is the one real big downside, it seems to me of this whole paradigm is that these things can get too hot and just blow up and that there's no bueno, right, so what can you do happen? Yeah, so.
We I mean like thanks to the technology that that is like a new layer I would say about understanding better these assets and using them better. I think that you know right now that are the right technology and also the artificial intelligence and machine learning capabilities to better understand how to again manage and optimize these assets.
That will be also.
You know, well we'll have we will have so much a broader adoption of these assets. And so this will also change the paradigm of how we think at energy in general and also the all the energy transition and everything is based on this because of course it's not only again the renewables, but you also need some other components to make this happen and also to sustain a much lower environmental impact, and so all of these will be much more necessary, especially technological side, for like making
this transition you know, a reality. And if you think out the grid right now is like you know configure for instance, you know and you have like you know, the traditional system in the western countries, like you have big plants where you produce the energy and then it's distributed to like you know, to the grid, to different facilities.
That is like off taking the energy right now with having you know, solar farms, wind farms and also you know BV panels in like domestic and business facility, so like for reastance on your rooftop or like you know, in some olar situations, you basically are changing this perspective and changing the paradigm of how the energy flows are
actually you know, pumping in energy into the grid. And so you need again capabilities technological capabilities to keep everything working and running smoothly and also to have additional benefit to all the users involved.
What's next? What are you guys working on now that we can expect in like a year or so. What's the next big push for you? Oh?
I cannot share that I would like to, but that's a good Yeah, that's pretty funny.
It is like yeah, yeah, I mean I can share.
It's like, you know, we are implementing day after day our models and you know, understanding what are the new needs of the customer. And I think the most important stuff this is also part of my job is like
getting ahead of the curve. So it's like what trends are going to be there in the next six, twelve, eighteen months and what we can do to address those in advanced So one of the things that we've always been able so is like understanding the trends before they happen and you know, getting something out before that moment so that we be leading.
Well, this is fantastic. And the company is Elektra. That are you based in Italy or you're just in Italy? It's a company based in Italy.
No, the company is based in Boston, Massachusetts, so the main edportter is there and then we also second European headquarter into in Italy.
Excellent, Well, that sounds going to be great, great, great, great information, great content. I love what you guys are doing. Very cool. Look these guys up, Giovanni Rossi of Elektra. We'll be right back. You're listening to Inside Analysis.
Welcome back to Inside Analysis. Here's your host, Eric Tavanaugh.
All right, folks back here on Inside Analysis. Your host here, Eric Kavanaugh, and I'm very pleased to have another guest with us today. Timothy Baines is the founder of a company called Compatio, doing really interesting stuff they've built a whole platform for you could say configure price quote. Really it's a platform for managing compatibility hence the name of
different products. So think electrical engineering or building materials, or anytime you need to bring a bunch of big pieces and parts together to create some whole like I think a car, for example, or any kind of complex machinery. You need to know are these parts compatible, do they work together? And then understand pricing and be able to figure out what to charge for these things. And they have a whole platform that does that, and they're focused on a couple of markets, but they can do all
kinds of different stuff. But Timothy, what really struck my interest here is you mentioned this taxonomy, and if you would, can you tell us what you mean by a taxonomy and how does that fit into what Campatio does.
Yeah, sure, Eric, So, first of all, I think it's important to understand that at the core, at the heart of our system is essentially a knowledge graph, and that knowledge graph sits on top of a decision taxonomy that within a given industry, let's say electrical switch gear for example, or industrial automation components, we have a very precise set of categories as far as what components are what, along with very high quality and very carefully engineered attributes on
the products as well as product catalogs for different manufacturers. So together the taxonomy which is the categories and the attributes, as well as the rules on what the attributes are allowed to be, what data type, what is the range or even what is the picklist of values, that comprises the taxonomy, and then that gets hooked up to catalogs of products, and together those feed into our larger sort of layers of logic that is essentially a knowledge graph.
Yeah.
And the reason why that's important for our audience out there listening is that when you have this graph and you have this ontology, you are cataloging all these myriad parts of which there can be millions of different ones, in such a way as to facilitate the alignment over time. Because if you were to try to manually put those into a just traditional relational database or something, first of all, it would take forever a day, and second of all, it would not be performed at all, it would be
very very slow, very painful. So by going this route of having a graph, a knowledge graph substrate with ontologies for particular industries or disciplines, if you will, you're really facilitating the end use of this whole system to allow people to compare and contrast and figure out do these parts work together or not. That's very very clever. And
you built the graph yourself, You popular this graph. Can you talk about some examples of how the graph and the ontologies facilitate what your end users want?
Yeah, absolutely, so if you think about well, if you think about any type of product that's available at scale, sort of on a large e commerce site, let's say all products need to have effective attribution and descriptions to power search. So that's kind of a fundamental use case. But when we get into more complex verticals, more complex product domains, where the attributes the specifications of the products need to be very very precise, you need an entirely
different level of quality and accuracy on those attributes. And the use cases that that powers are things like search. But you know search, including parametric search, right, so you can filter down and find exactly what you want, so any kind of discovery. But then the higher order logic that also that that enables in the domains where we operate the products are complicated. Finding the right product for
your specific application. Even just a single component can be non trivial, and oftentimes you may need to talk with an expert, let's say, that can help you figure out what the right product is. And then if you think about a product that is part of a larger system, an industrial automation system with a motor and a motor controller and a relay and so forth, at all need to go together to be able to identify those components number one that fit your application and the number two
go together. That's non trivial, and historically that task has been handled by experts, folks with ten twenty thirty forty years of experience in that industry, and you'd walk to the counter and a distributor, a dealer, a store, or they'd come to your plant and they'd have all the expertise that's needed to help you speck it out. But
things are going online. Obviously, digital commerce is expanding, and we also are seeing some fairly significant turnover in the industry in terms of the labor force, where a lot of the folks that came in twenty thirty forty years ago, they're retiring their baby boomers and they're rolling out and it's what our the companies that we're working with are telling is it's hard to find the right people, and it's hard to train them, and it's hard to retain them.
So we see really an emerging and a massive knowledge gap coming into as digital commresce becomes more more prominent. Combined with this problem of expertise, it makes it very difficult to do these kinds of things to find the right product fit your application, make sure they go together, and then price them and quote them and get them out the door. So we're addressing all of those use cases.
Yeah, it's really impressive. And when you talk about industrial engineering the machines that make machines, basically you have years on these things. Different models and things change your over year.
I mean, I'll use the metaphor of an automobile just because most people have an automobile or at least know someone who has an automobile, and depending upon the year of that make, you'll get different parts for different functions, like whether it's to run something in the engine or in the electrical system or whatever, and it really matters to get the right thing. I've been one of those do it yourselfer as I've bought some products I think it was going to fix, and it didn't fit the car.
I was like, oh man, because it was the wrong year. So it's really important to have the year right to make the model. And then all those attributes you talk about are basically edges on the graph, right, So you're able to capture all that and then maintain some history and thus facilitate finding that one specific part that you need to fix this machine.
Right, Yeah, that's right.
And you know, taking in a car and automobile as an example, I mean that data is generally very carefully maintained and curated by the manufacturer of that car. I mean even some of the legally required that they maintain that data for many types of things, for example liability reasons and so forth. But where it gets really interesting is where you, let's say, again keeping our focus on a car, you get out to what goes on the car or around the car, or attaches to the car,
or accessorizes the car. That data is not maintained by the manufacturer. So when you get into these multi brand situations, something that goes on or with something else, you're in a space that really nobody owns. Right now, where nobody owns that data what goes with what? Across brands, and in the industries where we operate electrical automation, building materials, and so forth, those problems are everywhere. Everywhere, folks are trying to figure out what goes with what, especially when
it's cross brand, but even within a brand, it's still challenging. So, yeah, the problem is huge, and the markets are huge. The companies in the industries that have these kinds of issues are huge. We're talking trillions of dollars worth of annual revenues in these industries and they all suffer from these problems.
Yeah, well you can just imagine. I mean, I've studied supply chain through SAP and some other companies, and it gets incredibly complicated because some of these companies will have hundreds of thousands, even millions of SKUs and so you know, just from a data management and storage perspective, that's difficult.
Then enabling search is difficult. Then enabling search for your partners and for your clients, which is also very important because you want them to be able to find the bits and pieces, the products, the component parts they need, and all of that is facilitated by this compatible platform right.
That's right.
Yeah, Search and discovery, search discovery, compatibility and configuration.
And then right once you can search and align and determine compatibility, then you get into things like pricing and configure price quote, which is another huge part of the industry. Knowing what you can charge for something, Knowing what the margins are that your competitors are putting out in the marketplace, that you're putting out there, finding the balance amongst all those things. I mean, that's a full time job for people.
And you still have to take guesses, right, I mean, you're going to figure out all your prices, You're going to configure and figure out Okay, I think I can get this price, and you send it out there and hopefully they bite, and then you go ahead and go through this, and you have to be able to go back and audit that and report on it and figure out where's the margin because you've got to make money, right, Everyone has to make money at the end of the day.
So you're helping facilitate these transactions. You're helping make sure that companies are getting the parts that they need to get, and you're also I think enabling more of that sharpening of the pencil basically, because once you know what the margins are, and once you know what the reliability of the providers are as well, then you can make these informed decisions. Then you can kind of understand where the margin is and that's kind of how you grow your business.
But folks, so Compatio co O m P A t io is Acompatio dot com. Is that right, Ai, Compatio dot Ai. Well, Timothy, congratulations on building this platform. This is a pretty seriously big deal and you're really greasing the tracks that I love that you've got a taxonomy or taxonomies ontologies basically and a knowledge graph. That's the way to do it. That's what our audience loves to hear about. So thanks for your time today. Look you spokes up online ladies and gentlemen. Compatio dot ai and
Timothy Bains will talk to you next time. You've been listening to Inside Analysis.
Welcome back to Inside Analysis. Here's your host, Eric Tavanaugh.
All right, folks back here on Inside Analysis. Your host here, Eric Haavadaugh, And today I'm with Abby Rass of a company called Hydraulics. It's hyd Elix, and Hydraulics is a very interesting company that is filling a significant gap that is enabling really powerful cloud content delivery networks and other companies to do real time analysis of observability at scale. And why does that matter. It matters for a lot of different reasons. It matters because these systems are very
complex these days. You can imagine Netflix, for example, the amount of data that's flying out all these various sources, all these various data centers to get to your TV so you can watch all these movies. That's a lot of data. So when you're trying to maintain these systems, you're looking for all kinds of things. And when you're trying to do that sort of industrial grade analytics on systems performance and management, well there are lots and lots
of metrics that you can look for these days. And obviously the Kubernetes movement has played deep into this whole space now spinning up amazing amounts of compute, federating compute basic, so really enabling this tremendous scalability of information systems of enterprise create information systems. But also it opened all sorts of Pandora's boxes on data and on what's actually happening under the covers, so to monitor that and to watch not just for bad guys, but also for performance issues,
anything that stumbles and starts causing trouble to users. You need to absorb tons and tons of data from a lot of different systems and be able to coalesce that and analyze it in fair fairly real time. And that's the kind of stuff that Hydraulics does. So they call themselves a bit of a streaming data lake company, but
also headless observability. And I love this headless concept. I'm hearing this more and more and I really like it because what it kind of implies is, look, we're giving you an engine and you can put whatever covering you want on it that suits your particular needs or your environment to use this powerful engine to get stuff done.
Right.
What do you think?
Yeah, I mean it's like mister potato head.
You can flip THEMP you can, you know, put on different hats, put on different eyes. It's whatever you want in the top. But you know the bum right, But no, it's it's essentially the same thing.
Right.
So with Hydraulics our software, anyone can plug in any type of dashboard.
It comes with pre built Graffona dashboards.
However, if somebody wants to keep their splunk instance, if somebody wants to keep any kind of front end dashboard that they have, that's totally fine. We have connectors so you can put any front end analytics dashboard on top of our back end and still get the benefits of less money, especially for storage, longer data retention, always hot data, for rapid queering and ingesting massive amounts of data at least to terabyte a month in real time.
Well yeah, and that real time nature is crucial. You were telling us a story before we hit the record button about a particular use case where a Black Friday real retailer was able to discover a problem as it was happening. Some kind of a hack was going on, found it, discover it, nullified it. Nobody even noticed on the front end, right.
Yeah, this was a large retailer global, It was during Black Friday week and there was a massive DDoS attack and it spanned eighty countries and three thousand IP addresses.
Wow.
Graphic Peak, the manager observability service that akama I built using our software at Hydraulics pinpointed this attack instantly, so the company was able to mitigate it in real time before any of their customers even noticed anything. Wow, So it really was instrumental for them to maintain business.
Yeah, that's amazing. And you think about what it takes to be able to ingest all that data, and that's a big key to what you're offering, right is number one, hyper scale ingestion of data. So you can tap into any number of systems, industrial grade systems, cloud infrastructure basically, and you are very good at pulling in data very quickly,
but then also enabling the analysis of that data. And that's all in hydraulics, right though you do also, as you suggest, you're headless, so you can work in conjunction with any number of analytics tools.
Right, that's absolutely true.
We actually just ran one of the largest sporting events in the country and at the peak viewing time, we were ingesting more than a petabyte of data an hour.
Wow, I'm sorry, not an hour, more than a petabyte.
We let me repeat that, at peak viewing time, we were ingesting more than a petabyte of data a day, Okay, which is an enormous I mean that's monster data, right, and a lot of a lot of times these legacy systems, because they were not built for the modern architecture of today will crash or slow.
Down when ingesting that amount of data.
But for us a petabite a day, nothing crashed, nothing slowed down.
We take it in in real time. We alert on.
Ingests, so we can show incidents in real time. You can query in three seconds, six seconds, super super fast, and then you can go fix the problem before anyone notices.
Yeah. Well, and you also you do some clever things on the storage side and some clever things on data reduction. Basically, so you really were purpose built for this kind of use case, right, I mean your executive suite they saw this coming, they saw a need for this, and then purpose built this engine to be able to fill this particular use case.
Right.
Yes, I mean it's two problems, right. It's the fact that the amount of data is growing so much each year it's supposed to increase by forty percent, that's when an analyst recently told us year over year. And also it's the cost. I mean, keeping data like this.
Costs a fortune. So we wanted to solve both of those problems.
The ability to keep all your data no matter how large the data set, and also make it affordable.
So that's what we're doing.
And the reason why, one of the biggest reasons with especially data retention, that we can make it affordable, well is that we have a twenty five to fifty compression ratio, so we make that huge data set very very small and easy to keep and managed.
Yeah, now that's clever. What's interesting is there are lots of different things that you can do from an architectural perspective to optimize for certain goals, for certain objectives, and this compression capability you have, that's why you're able to enable a much more cost effective storage layer, right.
Correct, And we have decoupled architecture, so compute and storage can be scaled up or scaled down separately.
Right, which is exactly what Snowflake did, right. I mean, I think they were the poster children for separating compute from storage. And everyone's like, oh, that's a pretty good idea. Let's go ahead and run with that, right.
Yeah.
In fact, in that largest football game, we were able to save the customer compute power because they didn't you know, when the peak viewing time dropped, they didn't need all that infrastructure, right, It scaled down really easily while still keeping the.
Storage well, you know, to myself here you talk about traffic Peak, which is a cool name, by the way, and that's that's the solution which Akamai built using your technology, right, which they're now selling to their clients as well. Is that correct?
Yes, we have so I leave partner marketing for Hydraulics, and we have very strong partnerships with Akamai and AWS and others, but our software can run on any cloud platform.
Akima is a huge.
Success story for us because they have been so supportive of the partnership from every level, from product sales to marketing. So when we launched in twenty twenty three traffic Peak, which is the manager observability service that Akamai built using our software, we had thirty three customers in the beginning.
In one year, that number shot up to more than three hundred and seventy customers.
That's great, and you know, it's a lot because the product speaks for itself, but also because we have such a collaborative relationship with Akamai and they're so supportive on all ends, and we're so supportive of them on all ends. So it's really been a successful partnership. And they just named us the North of America Qualified Compute Partner of the Year because it's great well.
And the thing to understand for from a broader business audience perspective is that we are seeing tremendous advances in all sorts of different spaces, and companies can come along like Hydraulics and build this engine purpose, build this engine for massive and jest for real time analysis. And when you do that, you're head and shoulders above some of the older legacy systems that just weren't designed to do
that kind of thing. And you know there are a lot, you know, so years ago I had doctor Michael Stonebreaker on the show. I don't know if you recognize the name, but he's the godfather of the modern database. Basically, he wrote the Postgress database like fifty years ago and has spun up several companies, including Vertica, which in its time in two thousand and five was all about real time analytics, was all about column oriented analytics. It was a new
crazy thing back then. But he had this one fun quote on a show one day where he said, the funny thing is that people should realize eighty percent of the code that's been written in the world should just be thrown away at this point was that it was written for older systems. And so if you think about spinning disc for example, versus solid state drive, well, it's a very different world if you're using SSDs in terms of that architecture and how fast you can move things.
And now there are all sorts of companies using the NVMe protocol to move data much more fast than it's ever been done before. And so I'm saying all this Abbey as a way to kind of tee you up, to explain to the audience, like this is you can't just upgrade your existing SQL server instance to get this kind of performance. I mean, this is a whole new engine that's been built for that specific purpose.
Right yeah, I mean that would be like creating mister potato Head out of wood.
I mean it's like mister potato head.
I forgot this can cancel culture.
But anyway, you know, these legacy systems were built in a certain way to handle data in a certain way, and what they didn't expect was for that data to get so big that it's literally like busting the seams.
Right right of these legacy systems.
And so that's why we need new observability services that can handle this modern architecture of all these micro services dispersed.
Everywhere that produce so much data.
I mean, just one application produces so much data because all of those.
Components that go into that app.
And so that's exactly why hydraulics exists today to solve for that.
Well, look these folks up online, folks, Abby Ross of Hydraulics. Very cool technology. You are listening to Inside Analysis.
We expected.
All right, folks back on the show here and now we're talking to Kenneth Stott. He is the field CTO from a company called Hasura, and they have a very interesting play for reporting, specifically for regulatory reporting, and they have a graph like approach to things. They use graph q well and it's sort of the last mile of data. And I really liked that concept and I that you threw out there that it's the last mile of data.
And you can understand why that's important, especially for regulatory reporting, because you give all these different information systems that people are using, whether you're in financial services or healthcare, and at the end of the day, you have to deliver these reports to people and if the numbers don't line up, that doesn't look good. That's when the auditor to say no bueno, you have to go back to the drawing board. So the question is where do you where do you
sort of reconcile all these different reports? And that's what Hisora does in this sort of supergraph layer. Right, tell us about that and how it works.
Sure, so I would describe it as a semantic layer. What that would What I mean by that is the individual teams that to create the data and own that data publish it in semantic terms into a data access layer, and then that data access layer validates all of the relationships that they that they've described within the information that they're publishing, so that you can get across all the teams one consistent view of your data in business terms.
And then that that same data access layer has a rich set of services to kind of discourage building new capabilities downstream, because oftentimes those new capabilities that they build downstream is where these problems start to arise, right sore, If they're cleaning data downstream, if they're making additional transformations,
this is often where it starts to go wrong. And the other thing I want to say about this last mile of data is this is the most consequential part, right, This is where people in the c suite make key decisions that have big impact on an organization, or where regulators pick up really meaningful differences that can drive you know, a lot of pain for an organization. Putting focus on this really pays off.
Yeah, that makes a lot of sense. And you know, I'm from the whole world of business intelligence and analytics and reporting, and you know, we came up with the data warehouse concept as a way of marshaling all the key data points into one relational database which we can then report from. So I get that. But to your point, a data warehouse isn't the only system of record. You have lots of other systems of record that these organizations
are focused on to shepherd through. And if you have this last mile, this semantic layer, or you're tracking these things, you can get a much better handle on it. And I like the way you say it's got some built in capability to discourage downstream modifications, right, Yeah.
Yeah, absolutely, data warehouses, you know, absolutely have to be part of the landscape going forward. There's a need to take data in and develop different analytics from it. But centralization has limits, and analytical data isn't the only thing that's required for these sorts of dashboards and reports. In this last mile of data, you need both operational data, you need analytical data, and so all of these things ultimately become the data products that flow into this smantic layer.
There are some amazing things you can do once you have visibility to all your data as it's in motion at this last mile. Another I've studied. I've studied then a lot of quantitative studies around where things go wrong. Oftentimes anomalies have to do with people combining data across disparate domains, and you can do things like anomaly detection at that level, right, and you can say, this really doesn't make sense the way you're combining this data to produce some output.
Yeah, that makes a whole heck of a lot of sense. And it is the reconciliation layer. It's the final proof really, and again you've sort of baked in here some of the formulae to be able to identify I mean, I won't it's not exactly like Master Data Management, but there is a flavor of MDN. There's a flavor of reconciling dispirit information systems. And I like too that you're not just looking at analytical data from a warehouse, you're actually
pulling in operational data. As a sort of reality check, right.
Yeah, yeah, absolutely, there are also upstream benefits, right, because what often happens in those data warehouses, those analytical environments is they start trying to serve a lot of different stakeholders and they start pulling in data from their peers, and you end up with a lot of cross sharing
of data in ways that become problematic. So I've also studied the cost of this sort of sort of ball of yarn that we generally create in these environments, and my take on it is about two thirds of data movement and data storage is wasted. If you could optimize, you could get rid of two thirds of that cost. And by the way, costs have been going up ten percent per year for the last three years, they're projected to go up again. And if you look at data
maturity statistics, they're flat. So we're spending all this money, right, but it's not getting that much better.
You hit the nail on the head. And it's funny you would say that because quite literally, DM Radio. Let's see, we turned seventeen years old yesterday. That's how long we've been doing these shows, and in the earliest days in the first year, I was trying to wrap my head around all the ETL that's being done to fill the data warehouse and thinking to myself, this is crazy. You guys are clearly moving the same data multiple times. You're probably not even using eighty percent of the data that
you're moving. Why are you moving it all? And it's because of how we got here. It's because way back thirty years ago, when we were beginning to really mature data warehousing architectures, people were just catching on and you're like, oh, well, I want this data in the warehouse. I want that
data in the warehouse. So you get these batch windows that stack up and it's almost like an archaeological dig now, right, you have just layers and layers and layers of data movement, and the question becomes, what do we really need to do? And what you're saying is that with this supergraph layer that you've got managing semantics, if people work with you appropriately, you can solve problems upstream, meaning you don't have to move as much data now, and you can also solve
a lot of problems downstream. And that's what goes to the executives to make the big decisions, right.
Yeah, So the idea is that the upstreams focus i would say, on core data sets. These aren't this is information I own. I own the definition combine. You know, both the tech and the business teams together. They don't get into serving sort of ad hoc requests because somebody randomly connected with them.
Right.
What happens then is the supergraph layer is sort of the place where those ad hoc requests get satisfied. And that's why the upstreams stop all this sort of duplicate, you know, sharing of data because they're very focused on just what they need to do and the downstreams. The downstreams get their needs satisfied more quickly because they're able to see the data, they're able to composite it into their narrow use cases for their needs.
Mm hmm.
We got about two minutes, a little over two minutes left here. You had mentioned before the call that you are containerized, right, like, so you run in a Kubernetes environment. Is that correct?
Yeah? So our particular product works in all the major cloud vendors of course, and you can do it as a SaaS offering, but it also works in a Kubernetes cluster, you can do it on prem and particularly think about healthcare and finance. There's a lot of consternation around public cloud and so forth, very important that you can do on prem when people have serious security concerns.
Yeah. Well, I mean I have to say I think that this focus on the last mile is really crucial, and I love that you're also helping upstream by satisfying a lot of these ad hoc style queries. I mean, the bottom line is that once you get analytics in an organization, people are going to want to know things, and they're going to want to run queries and they're
going to want to play with stuff. And the better you can shepherd those processes, the more governance you get, the more quality you get at the end of the day, and the low and you manage the cost as well. Right closing thoughts from you, go ahead?
Yeah, absolutely, I mean I couldn't agree more. The benefits are amazing. I'll say one last thing. I don't think it's the last thing you do. You don't get your house in order and then put this layer on top. Put this layer on top for current states so you can see what the heck is going on then fix it surgically.
Yeah, that's an excellent point as well. I mean it's you're talking about data observability, right. Everyone's into observability these days, into systems and you know, and tracers and logs and all these things. And I think communities is a part of that too, because it just opened up this whole can of worms about how to distribute process and everyone's like, whoa, now, hold on, how do we manage that? We how do we understand what it's doing because it's doing things very
very quickly. So but this is a great focus area and it's a layer of abstraction, as you suggest, the last mile, and I like that we say, understand what's happening now and then come up with the plant if things both upstream and downstream. But look this, gentleman up online Kenneth dot of Hasura h A s U r A correct. All right, folks, you are listening to Inside Analysis.
Now.
You can listen to k c AA radio anytime on your smartphone device call seven two oh A three five three oh nine nine seven two oh a three five three oh nine nine k c A eight.
Peur Life's much better, So download the app in your smart device today. Listen everywhere and anywhere, whether you're in southern California, Texas for sailing on the Gulf of Mexico. Life Sabreeze with KCAA download the app in your smart device today.
Ah B.
Yesterday in the basic.
K c A eight.
What is your plan for your beneficiaries to manage your final expenses when you pass away life? Insurance annuities, bank accounts, destment accounts all required defertivity which takes ten days based on the national average, which means no money's immediately available and this causes stress and arguments. Simple solution the beneficiary liquidity clan use money you already have no need to come up with additional funds. The funds grow tax defer
and pass tax free to your name beneficiaries. The death benefit is paid out in twenty four to forty eight hours out a deficitary.
Nerdy money without.
A definitive call us at one eight hundred three zero six fifty eighty six.
To Hebo T Club's original pure powder to RCO super Ta comes from the only tree in the world that fungus does not grow on. As a result, it naturally has antifungal anti infection, anti viral, antibacterial, anti inflammation, and anti parasite properties. So the tea is great for healthy people because it helps build the immune system, and it can be truly miraculous for someone fighting a potentially life threatening disease due to an infection, diabetes, or cancer. The
tea is also organic and naturally caffeine free. A one pound package of tea is forty nine to ninety five, which includes shipping. To order, please visit to Hebotea Club dot com. T hebo is spelled tea like tom, a, h ee b like boy o. They continue with the word t and then the word club. The complete website is to hebot club dot com or call us at eight one eight six one zero eight zero eight eight Monday through Saturday nine am to five pm California time.
That's eight one eight six one zero eight zero eight eight to Hebot club dot com.
Tune into the Faran Doozier show us you Mark the Place in Time the soundtrack to Life Sunday nights at eight pm on KCAA Radio, playing the hottest hits and the coolest conversations. Sunday nights at apm on the ferand Dozier Show with an array of music, talk, sports, community outreach, and veteran resources. With the hits from the sixties, seventies, eighties, nineties,
and today's hits. The Farran Dozier Show on KCAA Radio on all available streaming platforms and on a six point five m and ten fifty Am The ferand Dozier Show on KCAA Radio.
This important, time sensitive message is brought to you by this station's generous sponsor, George Let's Field Associates, who has important Medicare information for all current and future Medicare recipients about some big changes happening. Medicare Clarified. Medicare is a nonprofit consumer service organization.
It's more important than ever to review your Medicare plan for twenty twenty five from October fifteenth through December seventh to find out if you're in the right plan for you. People are calling nine five one seven six nine zero zero zero five nine five one seven six nine zero zero zero five. A popular and local Medicare plan is improving. Others are raising copays and adding deductibles, biggest changes in the Medicare drug program in fifteen years.
We thank George Letzfield and lets Field Insurance for their generous support of this radio station.
Here's the KCAA community calendar for March. On most Friday nights, it's open mic and karaoke in Riverside at Urrel Brewing Company, located on Chicago Avenue beginning at seven pm, So grab the mic and have some fun. In downtown Redlands on Saturday mornings from nine to one pm, it's the Downtown Morning Market located between six and eighth Street. On Saturday mornings, in the Riverside Main Library on Mission Avenue, it's Family
Storytime every Saturday morning at nine thirty. For car enthusiasts, here are a couple of options to enjoy cars and coffee in downtown Clarmat on the first Saturday of the month, the Claremont Car Guys and Gals. They meet in the village for coffee and cars at six thirty am. I hope you're an early riser. In Corona, it's more cars and coffee on Saturday mornings from seven am to nine am at the IHOP on Ward Low Road. All years makes and models are welcome. This weekly meetup allows you
to enjoy some beautiful cars. Looking to do some household hazardous clean up in Royale, so they offer a hazardous waste drop off on March first, March fourteenth, fifteenth, and June thirteenth from eight am till noon. The drop off site is located at two forty six South Willow And for those of you who love Pokemon in Redlands, there is a board game Teradise. Every Friday night at six thirty pm is their weekly Pokemon tournament. All skill levels are welcome, but you do need to have a full
understanding of the rules and how to play. Located at one oh nine East State Street in Suite A in Redlands, and that's the latest for the KCAA community calendar. I'm Lillian Vasquez on KCAA ten fifty am and one oh six point five FM.
Located in the heart of San Bernardino, California, the Teamsters Local nineteen thirty two Training Center is designed to train workers for high demand, good paying jobs and various industries throughout the lane An Empire. If you want a pathway to a high paying job and the respect that comes with a union contract, visit nineteen thirty two Trainingcenter dot org to enroll today. That's nineteen thirty two Trainingcenter dot Org.
NBC News Radio. I'm Chris Ragio. House Republicans are unveiling a short term funding bill to keep the government running through September.
Details now from Lisa Carton.
The six month plan includes cuts to non defense programs and the IRS. The ninety nine page bill is in the form of a continuing resolution known as a CR, which Democrats strongly oppose. House leaders say the bill has White House support ahead of Friday shutdown deadline. Speaker Mike Johnson said this week he thinks Republicans will be able to pass it along party lines, with a floor vote expected as soon as Tuesday.
The Trump administration is cutting four hundred million dollars in grants for Columbia University over pro Palestinian pro tests. Education Secretary Linda McMahon defended the move, citing anti semitism at New York City's Ivy League University. She said Jewish students have faced relentless violence, intimidation, and harassment following the deadly Hamas October seventh, twenty twenty three attack on Israel. Since last spring, the school has been a hotbed of pro
Palestinian protests, which have often turned violent. An unusual standoff in London today as a man carrying a Palestinian flag has been perched on a ledge on Big Ben's Elizabeth.
Tower for over ten hours.
Metropolitan Police say they've received a report about the unidentified man this morning, shortly before seven thirty London time. The BBC reports the man posted a video of his climb up and said he's protesting against police repression and state violence. Area roads and tours of Parliament have been canceled as emergency crews and police negotiators try to talk them down.
It's unclear if the climber is directly connected with pro Palestinian protesters who were in the vicinity for u planned rally. The Vatican says Pope Francis has shown a gradual slight improvement over the last few days. Over the weekend, the Vatican said the Pope's oxygen exchange has improved, he doesn't have a fever, and blood contests remained stable. The eighty eight year old went into a Rome hospital with