Resourceful Defense: Cybersecurity Strategies on a Budget

[00:14] Paul O’Dell: Thanks, everybody, for joining. My name is Paul O’Dell. I’m with CPP Associates. We’re partnered very closely with Centripetal. Fab, do you want to introduce yourself?

[00:25] Fab Fusaro: Hi, Fab Fusaro, Field Engineering here at Centripetal. I’m excited to be here with CPP to talk about ways we can protect clients’ networks and what we’re already doing with that.

[00:39] Paul O’Dell: Excellent. Thanks, Fab. I want to spend a couple of minutes on CPP, in case you’re not familiar with us. Then I promise I’m not going to get in the way of the great content Fab has set up for you.

[00:58] Paul O’Dell: I’m going to keep the slides short and sweet. CPP Associates is a solution provider based in the Northeast. We do a lot of different things, but I’d like to boil it down to what we’re seeing and hearing from our clients today. We talk about things from cloud to the need for turnkey AI solutions and recovery technologies. But the reason we’re here today is security.

[01:38] Paul O’Dell: What I hear from customers day after day is fatigue. They’re exhausted by the number of tools and alerts. They’re trying to figure out a way to have a good life—not even a better life, just a good one—because they’re exhausted from the day-to-day grind of securing the environment.

[02:00] Paul O’Dell: A quick introduction to who we are: this is our brag sheet. When you’re talking to somebody and asking, “Who are these guys? Are they real?” we’re real. We’ve been around since 2008. We’ve been lucky enough to win a variety of awards over the years. You can research us if you choose. We’ve been at this for a while, and we have a lot of experience on our side.

[02:27] Paul O’Dell: I’m not going to read this entire slide about why customers work with us. The most important point is that we believe in providing value in a way that doesn’t feel like you’re getting nickel-and-dimed. So many companies that do what we do will complete a project and do what I call the “leave-behind”: they leave something out so they can come back.

[03:11] Paul O’Dell: My promise to you is that when we do a project, we will transfer everything we know to your team. If you choose to work with us through a different engagement or a managed service, we’re happy to do that. But we promise not to hold anything back. We’re going to do our best and give you everything we have.

[03:36] Paul O’Dell: Part of why we can do that is how we’re staffed. We staff two to one, technical to sales. That is the cornerstone of what we’re about. The idea is to overwhelm you with technical resources. Most of our competition does it the other way around: a lot of sales resources, then one or two technical resources. We flip that on its head. It allows us to be consistent from pre-sales to post-sales, dive deeper, and invest in our clients.

[04:17] Paul O’Dell: Our service offerings include cloud, artificial intelligence, security, and data protection. Our delivery models include a fully managed service, co-management—which a lot of our customers choose—or implementation and handoff. It’s your choice.

[04:41] Paul O’Dell: Sometimes pictures tell a thousand words. The castle-and-moat strategy was fantastic during the early Middle Ages. You build a big castle, make the walls high, and put a moat around it. You could drop oil on people or hit them with bows and arrows as they came up. It was an excellent strategy early on.

[05:11] Paul O’Dell: However, as time progressed, the cannonball was invented. Fast-forward further, and the rocket was invented. Quite frankly, the castle-and-moat strategy no longer worked.

[05:27] Paul O’Dell: Obviously, I’m playing with pictures here, but picture that strategy as the modern-day firewall and endpoint protection strategy. It’s not good enough anymore. It’s overwhelming. Think about the first slide on alert fatigue, exhaustion, and tool sprawl.

[05:50] Paul O’Dell: Fab is going to talk about putting a shield around your business. That’s what Centripetal does. It keeps the cannonballs out and knocks down the rockets, making for a more effective overall security strategy.

[06:09] Paul O’Dell: Enough from me. I’m going to stay on mute, but I’m around. If you want to ask a question at the end, please feel free. Fab, the floor is yours.

[06:18] Fab Fusaro: I love the analogy. Thanks for the pictures as well, Paul. Very cool stuff.

[06:27] Fab Fusaro: I have a few slides to share, and we’ll go through them quickly. The idea is to explain what we’re doing at Centripetal that’s fundamentally different from anything else currently on the market. This technology combines a set of patents and technologies that have been used separately before. We combine them into what we feel is a very effective approach to protecting your enterprise.

[07:03] Fab Fusaro: This is self-evident to anyone in the security space now. The castle-and-moat analogy Paul used is represented by this picture of a guy holding his head, trying to figure out which aspect of the security stack has failed this time. We’re currently in a detect-and-respond paradigm, which is causing what Paul referred to as alert fatigue.

[07:44] Fab Fusaro: We also have a shortage of people. That’s a big problem. Currently, we’re around four million people short. Without the human talent to get ahead of this, you’re going to leave windows and doors open in your enterprise, and somebody will get into the castle.

[08:12] Fab Fusaro: This slide brings up something interesting: as the cost of connectivity has dropped over the years, the cost of cybersecurity has increased. As the world has become more connected, more criminals have had the opportunity to connect. We’re in a situation where it’s easy to connect to the internet and hard to protect yourself.

[08:48] Fab Fusaro: If we look at the growth of the internet from a connectivity perspective—actual ports, servers, and so on—the attack surface has exploded. Early on, we were talking about tens of thousands of web servers. Now we’re on the order of billions. These are the dynamics creating the challenges we’re facing today.

[09:39] Fab Fusaro: These bubble charts are from a website called Information Is Beautiful. I don’t think anybody is surprised that the challenges are getting worse. We always hear about the big breaches, but we don’t hear as much about the small ones. If you look closely at these publicly announced breaches, you see tiny dots alongside larger bubbles. The tiny ones represent smaller organizations and smaller exfiltration sizes: fewer records.

[10:19] Fab Fusaro: The tools bad actors use to get into your network have become commoditized. Almost anybody can become a hacker if they put their mind to it. There are plenty of toolsets available. We’re seeing smaller organizations getting breached more and more, consistently increasing over the years. This exacerbates the cybersecurity talent shortage we discussed earlier.

[10:56] Fab Fusaro: I pulled a few articles to talk about the lack of accountability, part of which is regulatory. Most people may have heard of the National Public Data breach last year. Everybody I know has been getting annoying emails containing Social Security numbers, phone numbers, and so on from hackers saying, “I’ve got your data,” and trying to elicit a response.

[11:31] Fab Fusaro: A lot of that is a side effect of the NPD breach. They’re bankrupt now, but that was billions of records. Pretty much all our personally identifiable information is out there on the internet at this point.

[11:44] Fab Fusaro: Municipalities and K–12 schools have experienced an explosion in cost per data breach. They tend to be smaller organizations that aren’t as sophisticated in their cybersecurity practices, so they’re getting hacked. That’s creating a new focus on accountability within the data broker space and the public sector.

[12:23] Fab Fusaro: The bigger challenge overall is that cybersecurity is hard to quantify. If you look at your cybersecurity stack and say it’s nine out of ten effective or five out of ten effective, how do you count that? These are hard metrics to establish.

[12:48] Fab Fusaro: There have been attempts to create key performance indicators, or KPIs, and key risk indicators, or KRIs. ISACA and the insurance companies are interested in this. They’ve been trying, but it’s a moving target. The tactics constantly change. How do you measure the new approaches bad actors are using to get into your network? It’s difficult.

[13:26] Fab Fusaro: Let’s talk about AI and machine learning. I want to get a few definitions out of the way because I’m going to talk specifically about machine learning and the differences between these terms.

[13:44] Fab Fusaro: AI is the idea of using computer science and data to enable problem-solving in machines. Machine learning is a subset of that. It’s about data analytics and statistical analysis, using those to drive algorithms that help machines learn and adapt. Deep learning is a specific kind of machine learning technique.

[14:14] Fab Fusaro: Why bring that up? When we talk about AI and security, a lot of people think of ChatGPT and generative AI. Predictive AI is really what we’re talking about when we look at AI use in cybersecurity. We want to analyze patterns, find bad actors, and identify the techniques they’re using so we can stop them preemptively or proactively.

[14:44] Fab Fusaro: Generative AI does have functions in security, mostly on the bad actor side. They use it to craft particular attack vectors: an enticing phishing email or a spear-phishing attack targeting executives within a company, trying to trick them into creating a situation advantageous to the attacker.

[15:24] Fab Fusaro: Generative AI has its place, but when I talk about AI moving forward, I’m referring to predictive AI. Predictive AI has been used for over ten years to protect networks. It can also be used to help mitigate generative AI attacks. I wanted to get those definitions out of the way.

[15:50] Fab Fusaro: What do we mean when we talk about using predictive AI? Anomaly detection is one example. We’re looking for patterns of resource utilization, errors, or access patterns and establishing a baseline. We want to identify how the network should run in a standard day-to-day fashion and how it is actually running.

[16:19] Fab Fusaro: You hear about UEBA, user and entity behavior analytics. A fair number of tools and vendors work with that. Another aspect of predictive AI is analytics: predicting storage requirements or identifying patterns that may lead to failure.

[16:44] Fab Fusaro: Finally, we have clustering: recognizing specific traffic or behavior that deviates from a norm. That ties into anomaly detection. When we talk about machine learning, these are the algorithms and techniques we’re using in security.

[17:15] Fab Fusaro: This slide is interesting because I’m going to talk about threat intelligence shortly. I want to bring up data science. What we do at Centripetal is use threat intelligence at scale. Threat intelligence is essentially lists of risky elements on a network. Those could be IP addresses or URLs, such as a phishing link. We call these IOCs: indicators of compromise.

[17:55] Fab Fusaro: When you aggregate lots of indicators of compromise, that’s a data science effort. The reality of data science is that a lot of it is drudgery. You need machines, algorithms, and approaches to clean, process, and curate the data before it can be used.

[18:20] Fab Fusaro: The machine learning aspect depends on what we call data engineering. That is probably the most important aspect of doing effective machine learning for cybersecurity. It’s not trivial to take massive amounts of threat intelligence or IOCs and make them actionable. It’s difficult.

[18:56] Fab Fusaro: The point I’m making is: don’t try this at home. If you do, make sure you have enough resources—money, computing power, and smart data scientists—to do it.

[19:14] Fab Fusaro: Let’s talk about the fundamental problems in modern cybersecurity. These are probably the most common tools in cybersecurity stacks today. When analyses are done on breaches, there’s a firewall in place 100% of the time. We know firewalls alone are not enough.

[19:42] Fab Fusaro: The second and third most common elements present after a breach are endpoint detection and response agents, or EDRs, from well-known companies such as CrowdStrike and SentinelOne. In 90% of breaches, there’s endpoint protection in place. If there’s endpoint protection, there’s probably antivirus software bundled inside that system.

[20:13] Fab Fusaro: We know the top tools in the industry aren’t able to fully protect enterprises. It’s not that these technologies are bad; bad actors have figured out how to get around the tools we’re using.

[20:36] Fab Fusaro: Let’s talk about cyber threat intelligence. I mentioned IOCs before, and I want to explain what they’re made up of. The first types are URLs and URIs. This is your classic phishing link: something you click, and the next thing you know, you’re installing software or your hard drive is encrypted. Now you’re in trouble.

[21:05] Fab Fusaro: A less precise indicator of compromise is a domain. You might go to a website that we know is bad, but your users might not know that. Domains are constantly changing. Creating a new domain is a common vector for bad actors to try to get in.

[21:29] Fab Fusaro: Homoglyphs are a classic example. I can take something like Facebook, replace the O’s in “book” with zeros, and get somebody to click on a link. These are examples involving base domains or fully qualified domain names.

[21:50] Fab Fusaro: Then we have what are called multidimensionals: an IP address and a port, or a socket. We also have IP addresses themselves. We see IP addresses trying to get into our network. We know those are bad actors, and we know which countries they’re coming from. Those are common IOCs.

[22:09] Fab Fusaro: Finally, we have geolocations, which are often tied to IP addresses. Once we see a set of IP addresses coming from a specific country, we may decide to block that country. I don’t think anybody in the States is doing a lot of business with North Korea, so North Korean IP addresses probably shouldn’t be on your network unless you have a solid business reason.

[22:48] Fab Fusaro: As I mentioned, these vary in fidelity. A URL is a very specific IOC. A geolocation is a very coarse IOC. If I block an entire country using a geo-blocking rule, that’s a hammer. I’m saying, “I don’t want any traffic from that country.” Blocking a URL is difficult if you don’t know exactly what you’re trying to block.

[23:18] Fab Fusaro: IOCs span this range of fidelity, and you need all of them to protect your network. The last type I want to discuss is IDS rules: intrusion detection signatures. They’re also called Suricata rules or Snort rules. These are patterns that can encompass all these types of IOCs.

[23:46] Fab Fusaro: IDS rules are important because this is how zero-day vulnerabilities are identified. Think of MOVEit or Log4j. When they’re published as a zero-day vulnerability, they’re typically published as a Suricata rule. You use a combination of IOCs to establish a pattern you want to scan for and keep out of your network.

[24:13] Fab Fusaro: That’s enough of the tutorial on cyber threat intelligence, but it’s important because it feeds into how we can use threat intelligence to protect your network.

[24:31] Fab Fusaro: We talked about how, after a breach, we analyze it and find specific tools that were circumvented. When we look further and find the actual cause of the breach, 95% of the time we know the IOC existed somewhere. That indicator of compromise—the ultimate reason for the breach—was known, but it wasn’t used.

[25:09] Fab Fusaro: That’s not just us saying it. Gartner talks about this, as does Webroot, a cyber threat intelligence provider. Gartner talks about 99% of vulnerabilities being known to IT staff. A lot of that has to do with patching regimens.

[25:20] Fab Fusaro: For example, I know my web infrastructure is running a particular version of software, such as Apache or NGINX. I haven’t upgraded or patched it to protect it. I don’t have time, or I can’t afford to take the systems out of commission, so I have to schedule the upgrade. It’s not necessarily negligence. Sometimes it’s not practical to get to that maintenance window.

[26:07] Fab Fusaro: Webroot talks about locations. Ninety percent of breaches occur from locations where we know there are bad actors. I don’t want to pick on North Korea too much, but places such as Russia, China, and North Korea have dedicated advanced persistent threat actors trying to hack in and access data from our enterprises.

[26:38] Fab Fusaro: We know the tactics and locations, but we’re still getting hacked. Why? That 95% number is powerful. This is another bubble chart I brought up as an example. If I could take 95% of those bubbles away, the chart would look like this. That would be a much better state for our industry if we could prevent the majority of breaches. There is a way to do this using threat intelligence.

[27:27] Fab Fusaro: There are quite a few threat intelligence providers spread across the internet. They have different telemetry. Many have sophisticated honeypot networks monitoring hackers who constantly try to get into their infrastructure. They take that data and report it.

[27:53] Fab Fusaro: This is the threat intelligence ecosystem we need to draw from. It’s where those 95% of IOCs are contained. This list isn’t complete, but I’m pointing out some of the top companies providing commercial threat intelligence: IBM X-Force, Recorded Future, Sophos, GreyNoise, and Webroot. These companies are very good at tracking IOCs. The challenge is that their data often isn’t used. Let’s talk about why.

[28:33] Fab Fusaro: I call this the rainbow slide because I use different colors for different vertical markets: banking, healthcare, and higher education. Most organizations connect to the internet with a high-speed data pipe and a SIEM on the back end. The disks on the far right represent audit trails.

[29:07] Fab Fusaro: A SIEM, or security information and event management system, typically captures network activity. This is about audit trails and cybersecurity insurance. If you have cybersecurity insurance, providers may request a tracking system, such as a SIEM, to reduce your premium costs. This could also be an EDR; EDR tools often have their own built-in SIEMs. The idea is to have an audit trail of what’s going on in your network.

[29:48] Fab Fusaro: The way threat intelligence is typically used is that bad actors come in and out of existence. They appear, try to hack, and disappear. They might try a different IP address. As they hop around the internet, threat intelligence providers observe them and sell the data back to enterprises through subscriptions.

[30:22] Fab Fusaro: If you get one feed, or one provider supplying several feeds, you only get a small percentage of the IOCs present on the internet at any particular time. Even with two subscriptions, you might have only 2% of the threat intelligence. This is why it isn’t being used at scale.

[30:49] Fab Fusaro: That limits the effectiveness of detect and respond. Security operations teams sit looking at the SIEM, trying to identify patterns. Sometimes they use machine learning to determine whether a pattern represents a risk to the organization. But with this limited use of threat intelligence, you won’t catch the majority of activity.

[31:18] Fab Fusaro: I have some numbers here that speak to that. The top line shows the probability that an indicator of compromise, such as an IP address, is known to only a single threat intelligence provider. Let’s stick with the North Korea example. The probability that an IP address is known to only one provider is about 82%.

[31:49] Fab Fusaro: For domains, it’s 94.26%, and for URLs, it’s 99.43%. You might identify a phishing link based on a pattern, but the chance of capturing it because it’s a known IOC is almost nil. This is why we’re consistently getting eaten alive by phishing and ransomware.

[32:25] Fab Fusaro: This analysis wasn’t done by Centripetal. It was done by Enterprise Strategy Group. They looked at a critical mass of threat intelligence across a hundred different providers.

[32:37] Fab Fusaro: What Centripetal does is aggregate and curate the largest commercially available dataset of cyber threat intelligence in the world. That includes commercial feeds, open-source feeds, and some government feeds. We buy up that threat intelligence and apply our AI curation technology.

[33:03] Fab Fusaro: If an IP address is seen across two providers, we can change the risk score because it’s been seen twice. If it’s been seen within the last five minutes, that means it’s a higher-risk IOC, and we can change the ranking.

[33:19] Fab Fusaro: We have a technology called RuleGATE, which is essentially a set of patents. It’s an appliance that sits on the network at Layer 1, outside the firewall. Notice I said Layer 1: it’s essentially a piece of wire.

[33:36] Fab Fusaro: We look at every packet coming in and out of the network and match those packets against this huge threat intelligence database—what we argue is about 99% of the known threat on the internet. If a packet matches an IOC, we generate an event and send it back into our SIEM.

[34:04] Fab Fusaro: On the front end of our system, we have AI analysts: machine learning algorithms that process those events and output to a human analyst. This ties into what Paul discussed about managed and co-managed services. It’s typically delivered as a service through CPP.

[34:29] Fab Fusaro: You have a dedicated Centripetal engineer assigned to track risky traffic coming in and out of your network. We can tell you which devices inside your network are compromised and which ports bad actors are scanning. It’s a dynamic process. I have a use case that goes into more detail.

[34:51] Fab Fusaro: Let’s go into the use case. Hopefully, this will bring home exactly what we do. We call this shift-left security, and the reason will become apparent.

[35:06] Fab Fusaro: I’m showing the typical detect-and-respond paradigm used in the industry. We have a security operations center, or SOC, with analysts looking at IOCs represented here as dead bodies. That silhouette next to the analyst represents traffic that has come into your network. It’s already gotten in, so the deed is done.

[35:40] Fab Fusaro: Now we’re asking, “Was there risk? Do I need to look more closely at a PC or server to see whether that network transaction or those packets created a risky condition?” That’s what we mean by traditional detect and respond.

[35:59] Fab Fusaro: The customer in this use case is a large healthcare provider on the East Coast. They were using threat intelligence from multiple vendors. We talked about EDR and firewall vendors. Most established firewall vendors will sell you a threat intelligence feed.

[36:21] Fab Fusaro: Even with all the intelligence this customer consumed, they were seeing 63 million events per day they couldn’t identify. That generated 5.7 gigabytes of SIEM logs. You can imagine how overwhelmed the analysts reviewing that forensically were. This ties into Paul’s earlier point about alert fatigue and burnout. It’s a very real problem in cybersecurity.

[36:51] Fab Fusaro: We came in with RuleGATE. It’s an appliance at Layer 1, essentially a piece of wire. If you unplug this box, traffic continues to flow through it. It sits between the ISP’s router and your firewall, at different speeds.

[37:13] Fab Fusaro: We also request a link into your LAN on the back end so we can correlate outgoing and incoming events. When we say connect, we’re looking for a port span or port mirror so we can see the pre-NAT address: your private address range.

[37:34] Fab Fusaro: We see the event before it gets through the firewall. When it’s transformed into a public IP address, we can see it going out. We can stitch the events together and say, “Go look at that laptop, because it was trying to talk to North Korea.”

[37:52] Fab Fusaro: RuleGATE is now in your network. That critical mass of threat intelligence comes into our ACT system, the Advanced Cyber Threat database. That’s where we do AI curation. The data is pushed into the Global RuleGATE Manager, or GRM. The GRM pushes curated data into RuleGATE, which watches every packet and matches it against this massive threat intelligence database.

[38:28] Fab Fusaro: We’re moving what the SOC analysts and cybersecurity engineers do to the edge of the network. We’re looking at traffic as it comes in and goes out. That’s the first aspect of shift left: monitoring. We monitor for about a week.

[38:56] Fab Fusaro: After a week, we’ve accomplished three things. We’ve matched this massive intelligence database against your traffic, and we share that with you. We analyze the advanced threats. A lot of this is noise: somebody trying to get in but hopefully failing because you have good firewall rules.

[39:23] Fab Fusaro: We tell you which specific things are getting through, or which compromised devices are communicating with phishing sites, command-and-control sites, or botnets. Then we make shielding recommendations.

[39:40] Fab Fusaro: Once we apply the shielding recommendations, RuleGATE is no longer just listening. It’s dropping traffic identified as coming from bad actors. We continue to look at every packet to see whether new bad actors come in, but we’re now blocking the traffic we saw in the last seven to ten days.

[40:07] Fab Fusaro: It’s a dynamic process. Protection is now proactive. This is shift-left shielding. As new threat intelligence comes in from providers, we update that policy. We’re not shielding specific IP addresses; we’re shielding a specific bad actor. Even if that actor changes their IP address, RuleGATE is updated. This is dynamic protection.

[40:34] Fab Fusaro: In this case, the 63 million events the customer was seeing went down to 213,000 per day of unidentified traffic. The SIEM logs were reduced from 5.7 gigabytes to [unclear in source transcript: “11 hours”; verify the figure and unit against the recording]. That’s where we get this 99.6% reduction.

[40:51] Fab Fusaro: That ties into the 95% number we discussed for breaches that identified an IOC. We’ve been able to solve this problem by using threat intelligence outside the edge of the network.

[41:12] Fab Fusaro: The end state is a dramatic impact on the stack. We can even do things like zero-day mitigation. None of our customers were affected by [the vulnerability name is unclear in the source transcript; the speaker is interrupted].

[41:24] Paul O’Dell: You have a couple of questions in the chat.

[41:28] Fab Fusaro: Go ahead, Paul. I’m almost done.

[41:28] Paul O’Dell: I’ll let you wrap up, and then we’ll get to the questions.

[41:31] Fab Fusaro: Let me finish. Zero-day mitigation, a massive impact on the network and assets, and a big impact on the security staff. We’re able to alleviate alert fatigue. Paul, go ahead.

[41:46] Paul O’Dell: What questions do we have? Do you support cloud infrastructure?

[41:52] Fab Fusaro: Yes. RuleGATE can be deployed in AWS and Azure using gateway load balancers. We roll that out with a Terraform script and can protect your infrastructure as a service.

[42:07] Question from the chat: What about decrypting TLS/SSL?

[42:11] Fab Fusaro: We can decrypt TLS and SSL. We don’t do it by default. Typically, people are already doing that on their firewall. If we see something risky that’s encrypted, we’ll tell you and say, “We should probably decrypt this.” But it’s up to the customer to approve that. Your payload is none of our business.

[42:35] Fab Fusaro: If you’re okay with us decrypting it, we’ll do that. The packets stay on the device. You have control of the device. We can only look at it with you. If you want to download the packets and send them to us, that’s up to you.

[42:48] Fab Fusaro: This technology can do TLS/SSL decryption, but it’s all under the customer’s control. The customer keeps custody of their data.

[43:02] Fab Fusaro: I want to show some reports. This is a typical cadence where the first week is shielding. Over the next three weeks, we eventually get to 99%. That was also in the cloud, which ties into the cloud question. It’s powerful to run this in the cloud.

[43:20] Fab Fusaro: I did this trial when I was working at Facebook a few years ago. That’s how I joined the company: I was so impressed with what they do. There is a 30-day no-cost trial, and you can talk to Paul and the CPP team about that. You don’t have to believe everything I said here; you can experience it yourself. Reach out to your CPP team. That’s all I have.

[43:46] Paul O’Dell: Thanks, Fab, and thanks, everybody, for hanging in there. What a great story. The trial is very eye-opening. The typical response is, “Oh my God.”

[44:00] Paul O’Dell: Fab, I like what you usually say: you see things you can’t unsee. Now you have to take action. As good as many modern devices are, think about the castle-and-moat strategy and the ring slide Fab showed. One hundred percent of the incidents had a firewall, and 90% had endpoint protection. Things are happening, and they’re probably happening on your network. We definitely want to help.

[44:32] Fab Fusaro: For sure. Everybody should at least kick the tires on this and get what I call a free MRI of your network. You’ll see exactly what’s getting through and what isn’t working. Once you see it, you can’t unsee it, and you won’t want to give it up. That’s what we’re banking on.

[44:57] Paul O’Dell: Thank you very much, Fab. Thank you, everybody, for joining. Please reach out to your sales team or contact Fab and me directly. We’d be happy to help. Thanks, everyone.

[45:08] Fab Fusaro: Thank you. See you, Paul.

[45:10] Paul O’Dell: See you, Fab. Take care.