EcoCast By ActualTech Media Guide to Cloud Tools: The Products and Services Critical to Cloud Migration, Security, Sustainability, and Success

  • Tiger Technology
  • Wiz
Guide to Cloud Tools: The Products and Services Critical to Cloud Migration, Security, Sustainability, and Success

About This EcoCast

Navigating the complexities of modern cloud environments requires a clear understanding of the platforms, services, and architectures driving modern enterprise infrastructure.

This EcoCast provides a practical overview of innovative platforms designed to streamline workload migration, fortify data security, and optimize operational efficiency. Participants will explore cutting-edge tools that reduce migration risks, automate compliance tracking, and integrate eco-friendly computing practices into core cloud operations.

Designed for technical decision-makers, this session focuses on practical solution evaluation, highlighting emerging products that maintain high system availability. Learn how leading organizations leverage next-generation cloud services to eliminate vendor lock-in, control operational expenditure, and build resilient digital environments.

Gain the actionable insights needed to select, deploy, and manage high-performing tools that align cloud performance directly with long-term strategic goals.

Why You Should Join:

  1. Discover next-generation platforms to streamline complex enterprise workload migrations
  2. Evaluate innovative tools designed to strengthen cloud security and compliance
  3. Identify cutting-edge software solutions that maximize long-term operational efficiency

Agenda & Speakers

  1. Nikolay Fesliyski
    Nikolay Fesliyski Product Manager Tiger Technology
  2. Tal Moriah
    Tal Moriah Product Marketing Manager Wiz
Full Transcript Keith Ward Hello, everyone. Happy Tuesday, and welcome to the actual tech media eco cast. Today's topic couldn't possibly hotter… Be hotter or more relevant to your work. It is called Guide to Cloud Tools, the products and services critical to Cloud, my… Cloud migration, security, sustainability, and success. I wanna start by saying a big thank you for joining us on the Ecocast today. We are going to be joined by two heavy hitters in the industry. They are Tiger Technology and Wiz, and we are thrilled to have them on today. I am Keith. For those of you who don't know, I'm a webinar moderator with Actual Tech Media. We are a future b two b company, and I am totally psyched to be your moderator for the EcoCast today. This is going to be a fabulous event, and and I'm gonna start by covering some housekeeping items you need to know. We're gonna begin as we always do. In the chat tab on the far right hand side of your screen, this is our virtual Starbucks where we hang out during the event. We talk amongst ourselves. We share stories, emojis, anything we are excited about. We get to know our community a little bit better as well. That is one of the main advantages to that chat tab. So what I'm gonna do is call out some of you who are on right now and where you are from. So Steven says hola. Vin is in Central Connecticut. Good morning to you, Vin. Tom is in Wisconsin. Steve in Wisconsin also. Jordan is in Colorado. Jennifer is in Dallas. Matthew says good morning all. Good morning to you. Matthew from Scully, California. I assume that is not a real town, but I guess you never know. Alright. Andrew is in Northern Illinois in Illinois. Martin is in Detroit. Ben in Arkansas. Nick is in rainy Seattle. Sorry about that, Nick. But if you're gonna be in Seattle, you gotta expect that. Right? That is what I love seeing. Please, everyone, let us know who you are and where you are dialing in from or logging in from more likely wherever that is. It's great to see all of you here. Now the other thing you can do with chat is get technical help. So if anything goes wrong, if you can't hear something, if you can't see something, if anything goes sideways, first give your browser a quick refresh or two. Normally, that clears out all of the issues. If that doesn't fix it, my spectacular colleague Riley is standing by to help you out. So hit her up in chat if you need help with any technical issues. Okay. Two doors down from chat is the q and a tab, and I bet you can't imagine what you do over there. That's right. You ask questions and you get answers, hopefully. So q and a on the right, chat on the left. Try to keep those straight if you can. It just keeps things a little more organized for us. So in the middle of those two is the docs tab. There is some great content from our sponsors that goes along with the webinar. Click around and explore while the event goes on. Okay. You have also got closed captions for those of you who are not native English speakers. So you go to the bottom right of the screen, and there is a drop down there where you can pick among the many languages we have. Please take advantages if if you need it. And as always, for an actual Tech Media event, you are gonna get more than fabulous education, but a couple of you today are gonna get $300 Amazon gift cards. Now you do have to be in attendance during the entire live event to qualify for the prizes, and all winners must meet our terms and conditions. And with that, it is time to get this show on the road. We start off today with an interview of cloud tools by the magnificent, the immortal, Scott Becker. Now you may not know this about Scott, but he was the original subject of that new documentary, You Can See Everything. But when he proved to be just too weird, the filmmakers pivoted instead to Elizabeth Holmes. True story? I don't know. I read it somewhere in the… On the Internet, so it must be true. Right? And with that, we're gonna let Scott take the reins here. Scott, my friend, you are on. Video Thanks, Keith, for that reliably ridiculous introduction, and hello, everybody. So a decade ago, the promise of the cloud was sold to all of us in terms that sounded almost utopian, infinite scale, frictionless elasticity, zero operational overhead, and a pay as you go financial model that would magically eliminate technical debt. That's my symbol for magic. Alright. So fast forward to today, look at almost any enterprise IT environment, and reality tells a markedly different story. Migration is no longer a one way, you know, lift and shift project with a clean end date. It's settled into a permanent complex hybrid architecture. Security teams are drowning in disjointed alerts while attack surfaces expand across multi cloud estates. Cloud expenditures are increasingly scrutinized, not just for cost, but for operational and carbon efficiency, and our data footprints are exploding at a rate that legacy storage concepts simply cannot accommodate. So today's event is titled the guide to cloud tools, and you're gonna see incredible technologies across migration security, you know, optimization, things like governance. But tools alone do not build resilience. Before we evaluate what to deploy, we must be brutally honest about why our cloud strategies stumble. So that's why I've got this matrix like title here, The Architecture of Reality, Unmasking the Friction Points in Migration, Contextual Security, and Operational Scale. Okay. So we need to examine the three foundational underlying friction points in modern cloud environments. I'm gonna start off with the data gravity paradox. This is the operational and physical friction of moving massive unstructured data. Then there's context block security. You know, that's this dangerous gap between finding isolated vulnerabilities and understanding actual exploitable risk. Then we've got the sustainability and sprawl, and I'll go ahead and say it, a crisis, the the hidden architectural costs that quietly bleed capital and compute. So let's start where every enterprise transformation actually lives and dies, and that's the data. So when organizations first map out their cloud migrations, application workloads usually take center stage. We containerize microservices, refactor code, and spin up Kubernetes clusters, but then we slam directly into the immovable object, which is hundreds of terabytes or often petabytes of unstructured file data. Surveillance footage, medical scans, CAD designs, media assets, compliance archives, you get the picture. So here's the problem. Traditional enterprise applications and user workflows were never designed to speak object storage. Legacy line of business applications expect sort of low latency file systems, local POSIX semantics, and predictable local drive mappings. Meanwhile, the cloud's greatest economic advantage lies in scalable distributed object stores. So this friction leads to two common architectural missteps. There's the the all or nothing trap where organizations attempt massive brute force data migrations. They spend months replicating petabytes over expensive pipelines, only to face extensive application downtime, broken file links, retraining hurdles for staff, and shocking egress bills the moment active workloads try to query that data from outside the bucket. Then there's the infinite hardware cycle. Terrified of, you know, cloud migration latency and and workflow disruptions, infrastructure teams continue refreshing expensive primary storage arrays on premises, kicking the can down the road while on prem storage runs out of physical capacity. So to bridge this divide, architecture has to decouple data location from data access. That operational challenge is maintaining a continuous single global namespace. High performance workloads or or, workflows, I should say, must continue hitting local NVMe or high speed primary tiers with zero change to application paths or use… User habits, while intelligent data management quietly operates underneath replicating blocks, tiering inactive files into low cost cloud, archive tiers, and leaving lightweight pointer stubs behind. When you solve the data gravity paradox with seamless transparent tiering, migration ceases to be a traumatic event. It transforms into an invisible, nondisruptive, continuous life cycle. So once our workloads and data reside in cloud ecosystems, the challenge changes immediately to protection. And this is where the industry's legacy security mindset has caused some deep dysfunction. So for years, security operations relied on two traditional mechanisms, agent sprawl and siloed scanning. If you wanted to protect a virtual machine, you installed an agent. If you wanted to check compliance, you ran a script against configuration APIs. If you wanted to check software bills of materials, you ran an image scanner in your CICD pipeline. So in a dynamic ephemeral cloud environment where containers exist for minutes, serverless functions run for seconds, and infrastructure is sort of spun up and torn down via Terraform, this fragmented approach collapses under its own weight. So you have things like operational blind spots. You can't install an OS level agent on managed databases, serverless compute, or third party cloud services. The moment an unauthorized or unmanaged shadow cloud asset appears, your agent based tools are completely blind to it. And there's alert fatigue and noise. Traditional configuration scanner will happily generate 50,000 alerts. It flags an unpatched CVE on a on a Ubuntu box and open port four four three, a disconfigured… A misconfigured as three bucket permissions policy and overprivileged IAM. Now you ask your security operations team which of those 50,000 alerts matters. Right now, they don't know because the alerts are disconnected data points. Okay. In the in the cloud, an adversary does not breach a system using a single vulnerability. They move through interconnected attack paths. So consider this real world scenario. Container has an unpatched high severity CVE. In isolation, that sounds alarming, but what if that container is buried behind an isolated private subnet with no Internet Internet route? Its real world risks, or or its real world risk is minimal. Now consider another container. It only has a medium severity of vulnerability. However, it's publicly exposed to the Internet via an open load balancer. It has clear text service account key sitting in memory, and that service account happens to possess administrative read write privileges to a production customer database. That's not a single vulnerability. That's a toxic combination. So the modern cloud environment demands graph based contextual analysis. Security platforms must look at infrastructure holistically, unifying configuration posture, vulnerability assessment, cloud identity, and entitlement management, and runtime signals without burdening engineering teams with hundreds of brittle agents. Disaster in the cloud doesn't happen because you lacked alerts. It happens because your real attack paths were buried underneath tens of thousands of meaningless warnings. Okay. And the third challenge is one that touches finance, DevOps, and corporate responsibility. That's sustainability and resource sprawl. When we first embraced the cloud, we were told elasticity meant sustainability because cloud providers run hyperscale data centers with lower power usage effectiveness ratings than than legacy server closets. The migration itself was hailed as an automatic green initiative. While hyperscaler efficiency is real, our internal architectural behavior often undermines it completely. We've got orphaned and and zombie resources, you know, across dev and staging accounts. Engineers spin up oversized compute instances. They attach non ephemeral block volumes, unattached IP addresses, and there's forgotten database snapshots. So when projects end, the instances are terminated, but the storage volumes or the backup snapshots quietly linger. And, you know, this can consume energy and generate monthly invoices in perpetuity. And there's overprovisioning as a crutch because engineering teams fear performance bottlenecks, compute workloads are routinely provisioned at three x or four x their actual baseline needs. In physical hardware, that was called buffer capacity in the cloud. That's a continuous energy and monetary waste that's running twenty four seven. Then storage inefficiencies, storing cold, unaccessed data on hot, high performance tiers burns unnecessary power and budget. So sustainability in cloud operations is not an abstract corporate ESG metric. It's the direct downstream consequence of operational discipline. A well architected cloud strategy treats efficiency, cost optimization, and sustainability as identical goals, and and identical in in priority. When you implement automated life cycle management, eliminate orphan resources, right size compute, and tier stagnant data to appropriate low power archive storage. Your carbon footprint and your infrastructure invoice shrink in perfect tandem. So as you spend today exploring that cutting edge, you know, of solutions that are that are featured in this showcase, I really encourage you to look at these technologies not as isolated purchases, but through the lens of architectural cohesion to build an environment that is secure, sustainable, and built for long term operational success, you wanna ground your roadmap in three core principles. First, unify, don't displace. Stop attempting traumatic rip and replace projects for enterprise data. Adopt solutions that bring cloud scale, intelligent tiering, and disaster recovery directly into existing application workflows transparently. Second, prioritize context over alerts. Demand complete agentless visibility across your cloud estate. Focus your engineering and security talent on severing exploitable attack paths and toxic combinations rather than chasing disconnected vulnerability tickets. And third, automate continuous governance, make efficiency, and life cycle management programmatic. If data management, access hygiene, and resource pruning require manual intervention, it simply will not happen at scale. The cloud is neither an automated panacea nor an insurmountable challenge. It's an amplifier of your underlying operational design. When you master the fundamentals of data mobility, contextual security, and operational efficiency, you unlock the agility and velocity the cloud promised from day one. Alright. Well, thanks for your time, everybody. Hope you enjoy the rest of the showcase. And, Keith, back to you. Keith Ward Alright. And thank you, Scott, for that awesome keynote. We never come out of one of your keynotes. Less informed, this is such great knowledge. It's so many important points you brought up there. I appreciate that very much as always, my man. So we now have our first poll question of the event up today right now, and you can see some of the results here. Let us know your time frame for adding new or updating existing IT solutions at your company. You can see we have got multiple, answers here, and you can see the percentages there so you can get a sense of where your colleagues stand in terms of, upgrades to the environment. In fact, that may be a reason you are here today to learn more because you are ready to do that. Alright. Now, while you are voting and continuing to vote, I'm gonna do a couple more quick shout outs here. Michael is in Pittsburgh. Hi to you, Michael. Allen, hello. And Andre is in Porter Ranch, California. Terry is in Kansas. Cheryl is in Nashville. Ally is in t o. I have no idea where that is. Could that be Toronto? I don't know. Or somewhere else. Alright. Akash is in South Africa. Fantastic. So glad you could be here. And Joe is in Oregon. Sorry to keep you waiting. I assure you, you didn't keep us waiting. Hey, everyone. Thank you so much for voting in that poll. Really important and interesting stuff for us. And you're gonna have more opportunities to vote today too. In the meantime, it is now, time to get to our presentations. Our first presentation in the EcoCast today comes from Tiger Technology, and presenting for Tiger is Nikolay Fazlitzky, product manager. We are gonna bring him on right now. And there he is. Hi, Nikolai. Welcome to the EcoCast. It is great to have you here. Nikolay Fesliyski Hello? Hello? Can you hear me? Keith Ward We can hear. you fine. You sound absolutely great. And with that, folks, a reminder to please put your questions for Nikolai in the q and a tab on the right. In the meantime, Nikolai, I am gonna get out of here and turn this over to you. I'll see you on the other side for some q and a. In the meantime, please take it away. Nikolay Fesliyski Yep. Hello, everyone. So I would like to, first present, our company, what we do. We are twenty twenty years in the business, and, we are… We have wide experience with, storage solutions, both hardware and software. We are pretty much a software solution company now. We manage more than a 100 petabytes of data, and we have a wide range of customers, all around the world. And we are, getting, ahead by, by closing round of investment just recently. So we partner with the the biggest players on the market, obviously, being in in the… In this business. We, we are recognized by AWS, a recognized partner, and also, verified partners for, mic… Microsoft Azure. We are actually a recommended solution in three major, solution areas. And, we have put just a few logos, here just to to showcase that we we have customers from medium medium entertainment to government and the really, really wide range of customers of ours. With that, what we do, we are here to present one of our, leading solutions called TigerBridge. This is a data management platform which stores your data in a open format. It's not a vendor lock solution. It, it enables you to take advantage of, AI, with your data to to empower it. So, we enable you to use any kind of, target, like, like I I showed the the hyperscalers, or you can build your private cloud or even you can use targets, as a as a on on prem targets. And, in the end, you end up with the with the… Whether hybrid cloud or or on prem solution of yours. And and we we are really, really flexible. The platform, is enforcing policies, which, enables you to place your data wherever you like it and, whether you wanna have a cost effect, whether you are chasing a cost efficiency or you're, trying to respond to to compliance, regulations, things like that. So, on top of that, we we have a lot of different, a lot of different, use cases, but the the main… The one which I'm looking to to, show you now, is the disaster recovery option. So you can… So let's let's talk about the different disaster recovery setups you you may wanna build. The first one, in the picture here, would, mean you have a complete mirror of your environment. That's, obviously, the most expensive one. It gives you, however, the shortest time to recover. It's, however, fairly unscalable. It's hard to scale. You you need to mirror your environment and have a second copy of, of your, setup in order to to switch over. Somewhere in the middle is is, smaller, standby side, the second option you might, choose. This, means you're you're running a limited version of your environment as a secondary, and, you use it only when, you need to, when for for whatever reason your, primary environment fails. The the limitation here, the limitation here is that that environment might be limited in terms of performance. It's not the best, use case for… If you're running one of those, high performing solutions. And the third and the most, economical option would be, no standby. So you can, have a setup which is, built from scratch. Well, built from scratch may not mean a lengthy process because, that could be fully virtual environment, for instance, and typically, it is. But you can also do it with the with the, with the hardware which is, sit… Sitting aside and waiting for you. Long story short, this would be your, longest, but, inex… Inexpensive, but longest option in terms of recovery time. What I wanna say is Tiger Beach can cover you for all those three options. So, and, here is, how a typical, setup would look like with the Tiger Bridge. So, on the left hand side, you have, your primary environment. On the right hand side, on the picture, you will see a disaster recovery side. This could be whether a virtual side. This could be a physical server running nearby, or it could be in a different location. As long as, this location is supported by your VPN connection, for instance, between the two sites. And, you can use any of the cloud providers, out there for for this purpose. So, Tiger Bridge would, would, would keep you covered on both sides. We'll synchronize the data and make sure that, you can switch to the secondary site as soon as you, anytime. I mean, as soon as you have a disaster on the primary side, you can… It's a really an easy process to to switch. So at this point, I I want to, do a quick demo. So, let me just move this around and hide the slides. Okay. And let me share my screen. Okay. Can you give me a sign if this is visible, Keith? It is? Yeah. It is. Okay. Alright. So what we look at is the Tiger Bridge, administrative con… Console. You, you can manage a couple of, different options from it. You can see the… Your license and your, options enabled for the license. You can see some, general settings. I'm not going through those, just, I wanna save some time and, focus on the main, presentation. And, in my case, I have just a single, source, and target pair here. This is a local folder on my PC, which, we… We'll be using for the tests. But you can you can have have as many, local for… Folders connected to a remote target. So what happens here is, the the… This local folder, in this case, is replicated to Azure. And, here on the right hand side, you can, quickly see my, account details. Of course, the key is hidden, but, this is what my container is and my my account and, account key. So in order to, simulate, a disaster, I will just remove, the source and target pair. So just quickly before that, I will show this is the folder on my PC. And, seeing those overlay icons, I don't know whether they're visible well for you. There we go. Those folders… Those overlay icons show the status of the files. Obviously, Snowflake would mean that, this folder contains, files in archive. Green would mean that the files in this folder are, local. And blue would mean that the files are partially, or fully in the cloud. And, what we have on on prem is, is a stop… What's called a stop file. This is a pointer to the… To your data. This doesn't mean that you cannot access the file. You can, and Tiger Bridge would automatically pull the data for you. So, going back to the, simulation, once I, disconnect this, I will delete this source. And right after that, this looks like a normal folder now. I don't, see the overlay icons. Tiger Bridge doesn't manage the folder anymore. But, we said that we will, simulate a disaster recovery scenario. So let's say this folder and all the data, and it disappears. Yep. I want to delete them all. Now it's… So just for the sake of the exercise, I will use the same folder. I will try to, I want to restore all my data. So what I'm going to do… And I will add a new source, and I'll point to this folder again as a source. And, here is a good picture showing the different options you have to connect to a different target. On this first, tab, you have, all the hyperscalers, the well known, cloud providers you can pick from. And you can switch to any, private cloud or even tapes. You can you can really use a wide variety of, targets to replicate to. In my case, I'm replicating to Azure. And, again, let me quickly take my login details just to show you how easy that is. So paste my account name, my account key. And once I click select container, if you remember, I had… I was using container one. That's where my data is right now. Right. And once I click apply… Yeah. There are a couple of options here you can pick from, but, that's not the exercise for. So, once I click apply, Tiger Bridge will, check the remote folder and, will recognize that there is existing data there. And there is, you're presented with, different options. And here is what, gives you a lot of flexibility. No action will mean that, no, TigerBridge will connect you to the remote folder, will will not, recover any of the data. You will be able to recover the data with the with the manual actions. But we want to, exercise, the semi or automatic recovery. Right? So import on demand would mean that, as soon as I, open a folder, it will populate the folder with the stub files. It will not retrieve the data. It will only, create stub files for me. It will look like… To the users, it will look like the data is there. This is particularly, convenient for very large datasets where you have, millions of files. So if if you if you know, even creating files, for a… On a file system, even if they're empty stub files, it will take hours. So that's an excellent option if you are dealing with a very large, dataset. I I recommend you using that. This means that, files will be created as soon as you try to open the folder. And, that that tool… You will not, require to wait for the whole process to finish in order to start working. Import all metadata would be the option to… For you to wait for all the step files to be created. That's quick process. Don't get me wrong. But, for millions of files that… Like I said, that's hours of, waiting until the file structure is there. Restore all data, of course, is the lengthy one. So this will pull all the the step… It will create the step files, but also will pull all the date… The content of the of the files. And link existing data will create just the, links to the files in the cloud. For this exercise, I wanna do import on demand. This is the option which would create files as soon as I try to open it. Okay. And one thing, you you need to do when when, you create a new source is to resume Tiger Ridge. It will be automatically post when you create a new source. So what happens with our folder? Since it was already, positioned in my, Windows Explorer, the the folder is, has recovered all the, root structure. And, sometimes, well, in this case, when I point a folder, it will it will, recursively check, what are the contents from the cloud and will will start creating the files for me. But creating the files, in fact, it will create stub files. So, at this point, you will notice that, the folder look a bit differently. I'm only, having stub files now because none of those files are retrieved from the from the cloud. That doesn't mean that I I cannot access them, and, usually, I use, video files. As you know, video files would take a lot of time. For instance, this file is, over 500 megabytes. So, if I do properties, you'll see that this file is, this is the file size, but it takes zero bytes on the on the disk. And this is a typical stub file. So if I try to open it, it will take a few seconds, but the file will start playing. Okay? So… And, somebody will say, well, this is because it's a prebuffering in the beginning. No. You can jump jump to any place. And, it's it's very, important to mention that, TigerBridge would monitor which blocks of the file you're trying to access and will pull only that data from the cloud. It will save, a lot of, money on on retrieval traffic, it will, it will look fairly instant as you can see. I was not waiting for all the data for the file to be, taken from Internet. So as you can see, we just cached 12 megabytes of the file. Okay? That works actually both ways. If you're a video editor and if you wanna edit this file, just a small piece of it, by pushing the… By saving the changes, only the changed blocks will be replicated to the cloud. So you're not getting the, large file uploaded through the network. So, with that, we have, quickly recovered from a disaster. We lost all our files, but but, as you can see, it takes really minutes for you to recover from that. And, we are actually looking to to make this even quicker for a very complex, for very complex setup. We are looking to, create a recovery console, which will, will recover very quickly, complex, Tiger Bridge configurations in in seconds. Okay? So with that, I'm done with my sharing. We can go to the last slide again. Okay. I think that's it. Okay. Alright. Well, I think we are ready to take questions if we have some. Keith Ward Okay. Nikolai, thank you so much. What a great presentation. That was absolutely awesome. I I think I speak for a lot of people on the event today when I say that I love in-depth demos, and that was super in-depth and useful. I love seeing products in action. And with that, before we get to q and a, a couple of things you do need to know about while we prep those questions. There is a new poll up right now. Please let Tiger Tech know what additional information you would like from them in the poll. You could see with the red dot right over it at this moment. So please vote in that while we're going through your questions, and you should also check out the information over in the doc section that Tiger has provided, for even more goodness. And with that, let's get into those questions. Alright. First question, today, Nikolai, is this from Adam. He says, we use cloud native AI and machine learning tools, but our data sits on premises. Since TigerBridge leaves local stub files but move the actual blocks to the cloud, can cloud based AI tools analyze that data directly in our cloud bucket, or does the data have to be rehydrated back to the on prem server first? Nikolay Fesliyski It's excellent question. Actually, you can, you can… If you're… If you need to go through all your data, the best thing is to do it in the cloud. You can do it because, data is stored in an open format. We don't, put the data in a vendor lock format. So you can directly do it in the cloud. I I I guess that answers the question. But in case your model is running on prem, that's not a problem for us. And, it… If it needs to read through all the data, Tiger Bridge can pull the data as soon as you're trying to access it. So it will be seamless process. You can do it both ways. Of course, the most the most, economical way would be to use the data from the cloud. Yeah. Keith Ward But, of course, options are always good to have. Nikolay Fesliyski Yep. Keith Ward Thank you. Okay. Next question for you today, Nicole. And, folks, remember to keep voting in the poll as we go on and through the q and a. Now next question. What is the maximum supported distance between the primary, and the disaster recovery site? Nikolay Fesliyski Yeah. Another excellent question. So, the replication between sites is actually done through the cloud. So you can put your secondary site Keith Ward Okay. Nikolay Fesliyski anywhere, really anywhere. As long as you can support the connection to the users to the secondary site, whether, VPN between the two sites or whether whether send them to work from home and connect to the secondary site, it can be anywhere. So it could be even on a different continent as long as, the workload, allows it. Yeah. Keith Ward That is really, good to hear. Okay. Next up for you. Can you replace your backup solution entirely and move to Tiger Bridge? Nikolay Fesliyski Yes. You can. So we support a wide range of policies to help you tune that, depending on your company, objectives. For instance, we we support versioning. We support, self delete, depending on the cloud you are using. So, you can fully replace your backup strategy with Tiger Bridge and, actually, continuous data protection, which we offer, is superior to any traditional backup you might have. Keith Ward Alright. Outstanding. Okay. Next up for you. And again, remember, vote in the poll and check out their resources over in the doc section. We have time for a couple of more questions here. It looks like, Nikolai. Appreciate you taking this time with us today. Next up, what happens if there is a short downtime on the secondary server? Nikolay Fesliyski Yeah. On the secondary server, we… Tagger reach when it's running, on the… On both servers, it it keeps them, synchronized all the time. So, imagine if if any of the servers, fall, falls down or fails, Tiger Beach will, will recover all the files as soon as the server is up. So you only need to bring it up. Tiger Ridge will do the rest to just, see what… Which are the changes which are missing from your copy of the data and to recover it… Keith Ward Okay. Nikolay Fesliyski To resync it. Keith Ward may… Nikolay Fesliyski Yeah. Keith Ward Yeah. That certainly makes it easy and much more simple. Okay. Next up, for you, Nikolai, is data replicated on schedule or as the file is created and updated? Nikolay Fesliyski Yeah. Well, data is, could be replicated, again, based on policies you defined. You can you can set it to be replicated, a second, at the second when the new data is, created or modified. Or you can even have a delayed replication. Yeah. For applications which are constantly saving files, we recommend you to, to use a postponed update so you are not making many, many versions of your file in the cloud. But if you want instant and, if you wanna shorten the the data which is, lost in terms of disaster, in case of disaster, we recommend a shorter time to… For replication. So long… Short answer, to the question is, it… It's up to you. Yeah. Keith Ward and that's… And that certainly makes sense, because every environment is different and the needs are different. You know, it doesn't… Some environments may have to to all be a hot tier, for example. Others, most of it can be cold or even on tape or something like that. So that is good to know. Okay. I think we have time for one more question for you today, and that is this. What are stub files that were generated when connected to the new source? Nikolay Fesliyski Yeah. Well, stop files, as as I explained, are, pointers. They're really zero byte files, which are supported by the, file system. They're pointers to, to the data, but, the the the purpose of STOP files is to to not occupy a space on your on prem infrastructure, but yet look like, carry the metadata of the file and look like the real file to the users. Tiger Ridge does all the work on the back end to just pull the data which you are looking to open, and that's a very, seamless process. So your application wouldn't, wouldn't, see the difference whether the file is local or remote. Yeah. So. that's what's the point. Keith Ward Yeah. Yeah. Just great stuff, Nikolai. We are running out of time now. We appreciate, all the time you spent with us in the presentation and in the q and a. And again, a big shout out, for the demo and for, introducing some of us, I'm sure, to Tiger Bridge. It looks like an absolutely, fascinating and great prod for to be trying out in the current age when, let's face it, all of our data seems to always be at risk. So, again, Nikolai, thank you for kicking us off, on the EcoCast with such a great presentation, and it was a pleasure to have you on with us today. Nikolay Fesliyski Yep. Thank you. Thanks, Keith Ward Thank. you. Alright. So we're gonna say goodbye to Nikolai there, and I'm gonna close-up the poll. And we are going to get to the prize drawing, the first one of the event today, one of two. And before we get there, you know what I like to see, and a bunch of you are already giving me the emojis, which is exactly what we want there. So with that, I am gonna give away this $300 Amazon gift card, and the winner is Daniel Judzio from Wyoming. Daniel, I saw you in the chat just a little while ago. Congratulations. He is in Powell, Wyoming. Alright, Daniel. You are the first winner, and that means we have got one more Amazon gift card to go, which I'm gonna give away after this final presentation today. And this final presentation does come from Wiz. Presenting for Wiz today is Tal Mariah, product marketing manager. Tal, it is so great to have you on the EcoCast. With that, I am gonna turn this over to you. Please take the floor. Video Hi, everyone. Thanks so much for joining today. My name is Tal, and I'm on the product marketing team at Wiz. I've been at Wiz for about a year, year and a half at this point, focusing specifically in the detection and response area. So I cover WizDefend, our product for detection response built for SecOps teams. And we know, in case you haven't heard it enough, that AI has changed the entire landscape for the whole security organization and specifically for Sec So at Wiz, we have been dealing a bunch with this new change. We've been doing a bunch of how SecOps actually expands in the AI era and changes their remit. And we've really come across a couple of different conclusions about how detection response needs to update and change for the AI era. So let's get started. So the new reality is here, AI has really changed the threat landscape. But what do we really mean when we say that? Because it's so easy to say that. It's so… It's easy for everyone to say whatever they want. But when we actually look at what's really changing, we anchor on three key points that AI is really changing in the threat landscape today. First, the threat model is changing. For the last, say, about twenty five years, cybersecurity was really built around deterministic software, fixed code paths, known logic flaws, hard coded rules. Right? And the threat detection response playbooks were really built for this world and based off that. Defenders knew that they could write a rule, build a playbook, baseline our behavior, and therefore understand deviations that are suspicious and probably malicious from this known behavior. In this new era, AI agents really break that foundation. They operate nondeterministically based on intent. That means that attackers actually don't need traditional vulnerabilities anymore. They don't need to exploit traditional vulnerabilities. An agent operating with a completely legitimate identity and authorized tools can actually be compromised at the intent level. Intent can be manipulated by adversaries through prompt injection, or it can drift over time through the agent's own reasoning. Both situations really pose a threat and can result in harmful actions, regardless of whether a human actor is involved or if the agent had intent drift on its own, right, which I think a lot of us have heard about recently. Secondly, we've seen that AI has drastically broadened the attack surface. Agents are running everywhere, locally on endpoints like I have on, you know, my machine with cloud code, across multi cloud environments, in SaaS applications, and inside the browser. These AI agents aren't just coding assistants anymore. They're running across the enterprise, automating the tasks of every team, and they run with broad permissions and tool access. Because in order to actually be useful to the enterprise, right, they need this access. They need that permission, or else they wouldn't really be that useful. However, that broad access really creates a wide potential blast radius for every single agent in the organization. An agent running locally on an endpoint may have access to execute changes in a cloud database or SaaS platform, extending the reach of a compromised agent further than maybe what teams would even know. Finally, AI has really collapsed our window to detect and respond to threats. If y'all are following the zero day clock as closely as I am over the last couple of days, we've seen this continue to go down and down and down, especially in the last few months. We know at the meantime, from vulnerability announcement to active exploitation has dropped significantly and only continues to shrink. This means that zero days are now weaponized within hours or sometimes even minutes of disclosure. Meaning that defenders who are… Who have to detect and respond must do so at true machine speed, at the speed that attackers can actually exploit these vulnerabilities and work on them. And the evidence is really there based on a lot of the attacks that we've seen, especially in recent months. The OpenAI Hugging Face incident is a great example of this. So for anyone who may have not heard about this incident, earlier this year, OpenAI was actually conducting an internal evaluation of one of its models. The evaluation that this model was doing was to asking the model to solve complex challenges on Exploit Gym, which is a benchmark designed to test model capabilities at turning vulnerabilities into functional exploits. To assess the model's true capability, the model was actually running with reduced safety refusals and in a sandbox environment without direct access to the Internet. As the model began to work through the challenges, Eric realized that actually computing and getting to the answers would take a really, really long time. Multiple sub agents then began communicating over improvised message boards that they literally improvised to, like, package management and on their own and actually realized that there may be a way and a different way to actually get the eval answers, to actually find the answer key itself. It did maybe the same thing that some of us thought about doing in college. Right? This actually marked a critical intent shift for the agents. They moved away from solving the problem internally to hunting for the answers externally. They determined that the answers… This answer key could be stored on Hugging Face, which is a popular platform for AI research and development. With the new goal of actually getting into Hugging Face, the model took the initiative to break out. It discovered a novel zero day in the package registry cache proxy, escalated its privileges, escaped the sandbox, and reached an Internet connected node. Once online, it changed its own credentials and zero days across Hugging Face infrastructure, was able to achieve remote code execution, and actually exfiltrated the solutions autonomously. And this incident is really notable. It's notable throughout the entire security community, but for us in particular at Wiz who focus on the detection and response world. Some things that we noticed. First, there's actually no human adversary involved. There's no traditional attacker here. This is an autonomous model pursuing a goal on its own and shifting its intent on its own. And the intent shift was actually observable. In retrospect, when researchers actually went back to investigate this incident, they could see from observing agent trajectories exactly where the intent went wrong and started drifting. Finally, latest models are really, really capable and really, really good at discovering and chaining zero days to reach their objectives, Whatever that objective may be, whether it's the initial object… Objective or an objective that actually drifted, they're really, really good at doing those really complex chaining. And this is the world in which organizations are being defended today. This is the world that detection and response teams need to secure. It's very, very different from the old world, and it requires an evolution. So in a world where agents actually act on intent with extremely broad permissions and at machine speed, what do teams actually need? Detection and response must evolve across three separate pillars to actually make teams ready for this world. First, we need to unify context from across every surface where agents run and interact, whether that's on endpoints, in the cloud, or in SaaS. We need visibility from code to cloud to run time to understand what agents are built to do and monitor what they're actually doing. And we need visibility where we've never had it before at the semantic layer of model prompts, reasoning, and output to give us that information in view into agent intent to help us understand what agents are doing, what their train of thought was, and and what those thoughts actually led to. Second, we need new detection capabilities built specifically for this new agentic threat model. We need detection engines that can analyze and reason on intent and detect and alert on intent shifts that lead to harmful actions. By combining context on an agent's capabilities, like its permissions and tool accesses, with an understanding of intent, we can anticipate and block before damage is done. And finally, we simply must become faster and more proactive. We need threat hunting, investigation, containment, and detection engineering work to work at actual machine speed so the SOC can consistently outpace attackers. We can't have human investigation be the bottleneck anymore in this new world where attackers can operate as fast as they do. But it's not all doom and gloom. Don't you guys worry. We actually believe at Wiz that defenders really have the advantage here. Attackers may have access to the same types of models that we do. They may have access to frontier models. They may have access to public and exclusive data. They have access to all the models that we use, and they may be able to work faster. But we believe that the defenders' advantage really lies in the context that they have across their environment and actually operationalize in that context with AI. We believe that this is a significant structural advantage that if used correctly, can really help defenders succeed and outpace attackers in this world. We all have context on our own environments and our own AI agents, while attackers have to actually guess from the outside looking it. And we actually own the complete inside map of all of our data. We also have context across the threat landscape at scale. We have continuous frontline threat data that we can ingest, and we can actually put this context to work properly if we have the proper AI system set up to actually traverse this data, understand it, and make it actionable. And when we actually look at some of these new attacks, we see where this data is actually really, really important in helping the vendors have and use that structural advantage. So as we see here, we see, I wouldn't call it common, but a relatively generic prompt injection attack and we can look at the anatomy. We see this prompt injection attack started with exposure with the attacker reaching the application through an internet facing ALB. We then see the prompt injection happening to the actual AI agent and we see that this coerces the agent to actually invoke a privileged MCP tool. We see that the agent is then forced into assuming a role of a new workload and then actually able to export sensitive data. The great news is defenders that have context across their environment have information across all of these levels. Right? You have the context about identity and access information. You have context about what's going on at the endpoint, and you have that telemetry. You have information about code, how developers build. You have information, and you actually have the code bases that are actually creating these agents. You have information and data on your cloud and on prem environment, which helps us define the blast radius of an attack. We have information about the AI applications and infrastructures that these agents run on. We have the actual application context itself, and we have information about the exposed services that these applications can actually be accessed through. All of this data in context is available and relatively easy to get for defenders as opposed to attackers who have to conduct guesswork and only need one way in to exploit. So what is our view on how detection response needs to develop for the agentic era? It's really those elements that we talked about. It's ensuring that there's visibility everywhere that agents actually run and interact, which covers endpoints, CICD systems, cloud, SaaS applications, and browsers. And it requires ingesting new types of telemetry to to actually enable that visibility and make sure that there are no blind spots in the environment. Then it requires actually putting all this context to work and getting that context across every layer where the agent operates, Those layers that we just went through. Right? So it's taking that visibility, layering on the context, and making it actionable. Next, we actually think the detection engines need to be expanded and specialized for these specific situations. Since agents act and are built so inherently differently from traditional applications, we think that new and specific detection engines are needed to actually understand, analyze, and be able to detect upon agentic intent. Finally, our second or third favorite word, machine speed, is required for containment and response actions. We believe that the best way to do this is through specialized trained AI agents that actually run and operate based on the context of your environment so they can actually take action, understand your environment, and be useful in the organization to help you guys operate at the speed that attackers are. And I will go into this a little bit on the agentic detection engine because this is somewhere where our team has spent a lot of time researching and really looking into. If you're interested, we did release a blog about this relatively recently on, building an agentic detection engine to actually analyze agent intent. And we're really, really passionate about this being one of the future areas to expand for detection response teams. We actually have a injected detection engine that focuses on normalizing model input and output telemetry across the entire cloud environment and turning these into enriched events that can help teams understand agent content better. The pipeline starts with model input and output log ingestion, kind of surfacing to surface immediately or slightly clearly malicious activity, and then actually using AI to fight AI and understand it with a small LLM analysis and a large LLM analysis to actually understand agent actions and be able to tag them as malicious, benign or security tests. The pipeline then scrubs and redacts information to ensure that it's able to actually create an enriched cloud event that can then be contextualized, whether it's on our security graph, poured it into a SIM for further action, understanding, and information. And, again, we think this is so important because agents operate so fundamentally differently in this world that a new way of detection response and more so detection on this side is really, really important to help teams ensure that they're alerting on actual malicious actions and not just alerting on every slightly suspicious situation that's going on because we know that that would increase a false positive rate significantly. So this is our approach to really helping teams ensure that they're able to detect accurately and at speed. If y'all are interested in learning more, we spend a lot of time, and by we, I mean me, writing our blogs, talking about our products, and sharing what we're thinking about in this new world. Please follow us whether it's on LinkedIn. You can keep up to date with our blogs or on any of our other social medias. Our team is releasing a lot of new information about this. And If you're interested in learning more, we have our contact email available. But thank you guys so much for joining. Keith Ward Okay. And thank you so much for presenting today, Tao. Just great stuff as always from Wiz. Again, one of the leading companies in this space. Now a reminder to everyone that the poll question is now up for Wiz. Also, let them know what additional information you would like from them, and they have got a handout over in the docs section as well for more information. Jorge says, great job, Tal. Very animated. Absolutely. I agree. She is extremely engaging and high energy. Reminds me of someone else that I know. Now, folks, before we get to the final prize giveaway of the event, I have got to ask you for a little bit of help and a little bit of feedback. Right now, Riley is opening a survey for you that we would love it if you could fill out. This is to help us help you. This is a way for you to tell us what you like and what you don't like about our events. So, tell us what, what would make you keep coming back, basically, and how we can do this better. This is super important for us so we can make these the kind of events that you, must attend because they are gonna be so valuable for you. So please do that now. And while you are while you are voting, I am going to go over really quickly, the handouts that we have got over in the doc section from our sponsors today. So Tiger Technology, our first, presenter has got a, PDF, a, datasheet on called seamless hybrid cloud data management with Tiger Bridge. And it says Tiger Bridge seamlessly extends your local storage to the cloud for cost savings, elasticity, AI processing, and data longevity, intelligently move data between on premises and the cloud, as well as across on premises systems for unified optimized storage. That is the Tiger Technology handout in the doc section. Now Wiz, who we just heard from, has got kind of a data page, which has all kinds of great information, including a one pager, everything you need to know on a single page to get up to speed on their technologies fast. And let's see. Here is what they say about securing agents one zero one. This is a report you can download also from, the docs section. It says AI agents are changing how work gets done. They take on tasks. Orchestrate tools that drive outcomes across environments as adoption grows. They offer powerful capabilities. At the same time, they introduce security challenges that require a clear understanding of how they work and how to secure them. That is exactly correct as well. So, some great stuff there for you to check out, and I'm gonna ask Riley to, close-up the survey again. Thank you so much for filling that out and helping us out that way. Appreciate it very much as always. And with that, it is now time for what you have been waiting for here. Another prize drawing, another $300 Amazon gift card for one of you out there in, attendee land. Thank you so much. Let's get some more emojis going here. Looking a little sparse thus far. Alright. Looking better. Looking better. Picking up. Alright, folks. There you go. We're getting there. We're at… Now we are there. Alright. I am totally, down now to give away this last $300 Amazon card. So the winner of that is Peter McGinty from Ohio. Alright. Congratulations to today's prize winners, Peter and Daniel. We are gonna be in touch ASAP about claiming your prizes. And folks, that's a wrap for the event today. I hope you enjoyed it and learned a lot. I did. I always do. I always love being on with you. You know that. Cannot wait to see you. In the next event, whenever that is going to be a big shout out to our sponsors today, Tiger Technology and Wiz. Without their support, do not forget none of us would be here. And with that folks, we're all done. We are wrapped, and I hope you have a great rest of your day.

Please provide citations & attributions for any information you generate from this content, thank you.