theCUBE + NYSE Wired: Robotics & AI Media Week | Tim Carmody, IPC Systems, Inc.

Clips
More from theCUBE + NYSE Wired: Robotics & AI Media Week

Tim Carmody

CTO

IPC

play_circle_outline Integration of AI in trading communications for real-time transcription and analysis

play_circle_outline Focus on trader productivity and workflow integration in trading environments

play_circle_outline Transition to cloud for TradeTech and development of AI-powered applications for traders

Info
Transcript

Tim Carmody, IPC Systems, Inc.

Tim Carmody

CTO IPC

In this insightful episode of theCUBE's coverage at NYSE Wired Robotics and Artificial Intelligence Media Week in New York City, we feature Tim Carmody, Chief Technology Officer at IPC Systems. The discussion explores the complex world of high-speed communications infrastructure that supports trading activities across financial markets.

Carmody, with their extensive knowledge and expertise as CTO of IPC Systems, outlines the critical role of trading communications technology in the financial sector. They explain the evolution and sophistication of tu... Read more

explore Keep Exploring

What new technological advancements are being developed for real-time transcription and data analysis in trading calls? add

What are the two big areas that a company is focused on in terms of trader productivity and workflow integration, and what partnership have they formed to address these areas? add

What new technologies and strategies is the company implementing in order to adapt to the changing industry landscape? add

bolt Powered by CUBE AI

Tim Carmody, IPC Systems, Inc.

search

>> Welcome back, everyone, to theCUBE's live coverage here in New York City. I'm John Furrier, your host of theCUBE. This is part of our East Coast access point, of course, we've got Palo Alto Studios connecting technology and Wall Street, bringing you the best of both worlds. And all the action's happening here, it's the Robotics Media Week, as well as AI Leaders. And Tim Carmody's here, CTO of IPC Systems, specializes in high-speed communications, which is a big part of all the infrastructure that's going on. Tim, great to have you on, appreciate you coming on.

Tim Carmody

>> Hey, John, great to be here.

>> So we were talking before you came on, you guys do a very specialty system called turret systems, or turret devices. I knew what it was, I looked it up, it made sense, it's networking. But explain, it's a very high-performance system, it's an integral part of some of these financial systems, what else? Explain what it is.

Tim Carmody

>> Yeah. And we really like to think of it as trading communications, writ large. One of the ways you interface with it is a traditional turret, but more frequently, it's now soft clients and mobile devices. And it's really anything that facilitates and optimizes trading ability, whether that's done through a voice connection or whether that's done through an electronic connection.

>> And so, I'm old enough to remember the old PBX days, and now digital communications, Wall Street, high-frequency trading was a big deal, people arbitraged the hell out of that, and now it's pretty standardized and monitored. What was it like on the evolution of the high-speed communications? What's the use case, multiple trades at the same time? Explain what's going on, and what are some of the technical things that have to happen?

Tim Carmody

>> Sure. So a lot of it is about always-on communications that can be blended together and you can have many connections at the same time. So think about the broker market, they'll be listening to prices from 32 different counterparties, let's say, hearing them all at the same time, and then being able to execute on them very quickly, and also being able to bring in a compliance officer or bring in another part of the deal. It's basically a voice communication network on a steroid in order to pull all these things together and just remove any friction from getting the deal done.

>> And speed is critical?

Tim Carmody

>> And speed. So in our world, historically, you talked about latch times, when you get on a dial tone, it's really slow. On our network and our systems, it's immediate. Now that technology has caught up a little bit, that's something we're all more used to, but we've been doing it for a long time before that.

>> So it's purpose-built infrastructure and hardware, right?

Tim Carmody

>> Very much so. Yeah, we do a lot of our own R&D and development and own a lot of patents, because it is purpose-built and it is unique to this industry.

>> What's some of the secret sauce, can you share? You've got patents so you're protected. What are some of the secret sauce?

Tim Carmody

>> Yeah. So a lot of it has to do with how you treat the call as it's happening and before it's happening, and how you're able to do a matrix to bridge all of these things together at the same time, and be able to do things like track all of the user behavior that happens at the time, whether somebody goes on mute, whether somebody goes on speaker, all of the impact of what happens from that point of view. And it's just a mass of data and quick functionality that happens that is unique to our environment and unique to trading overall.

>> There's an old expression we've been hearing a lot of, it happens a lot in the networking side of the industry and it's now happening in others, you've got AI for networking and networking for AI. It's kind of a hyperbole, it's kind of like the haymaker, people throw it around. But you guys have high-specialty comms, is there high-frequency comms for AI, is there AI for you? Because I can imagine, it's almost like a factory in my mind, maybe that's a bad word, but just thinking out loud, you have this purpose-built system, it's got this design, it's like a factory. Is there a digital twin angle here, does AI fit in?

Tim Carmody

>> Yeah. So one of the things that we've been very aggressive in doing with our customers is moving them to digitization voice anyway. You can hear the voice behind.

>> Yeah, I love the trade, it's going crazy back there. It's that time of year, Tesla's back on the board.

Tim Carmody

>> So now that we have the digitization of voice, now we can do a lot more with it in terms of making better workflow integration, all that, and a big area is now becoming AI. So if you look at what happens leading up to a voice call, what happens during a voice call, and then the compliance after the call, there's a groundswell of AI functionality that can be brought to that and that's what we're now doing in the industry. Just to focus for a second, which I think is more to the point of your question, during the call, being able to do real-time transcription, and then being able to pull in other data points, other information, being able to take an AI view of this, is something that is now real and available. One of the things we're coming out with in a couple of months is we'll call a head trader dashboard. So being able to listen in real-time to all the calls that are happening through AI, have sentiment analysis, have keyword searches that pull up different order books tied into all these other systems about what other trades have been done by this counterparty, all of that is a big part of where the world's going, and we're leading the way.

>> As CTO, I can imagine you're in these meetings, where it's like, okay, what's the overhead involved, because you're performance-focused. I imagine, pretty wild guess, but I think that you don't really want to disrupt the performance of the system. Of course, that's the case. But how do you guys look at things like overhead, because it has to be so fast, any AI coming in, you probably have a high hurdle in terms of the bar?

Tim Carmody

>> We do, and I think the North Star for us is that you can't interrupt the trading communications. We're a well-established, trusted vendor to the industry. Our goal, or our mission, is to keep everything stable and secure and sensitive, and then innovate on top of that. So when you talk about the AI overhead, one of the things we've done is offloaded that to an external system, or a separate system, from our voice trading communication. So even if there's a problem with transcription, it's not impacting the actual call, it's not impacting the ability to establish new calls or anything like that, it's separation of that .

>> That's a great point, that's a great point, that's a good thing to unpack. Let me ask you, because you're seeing a lot of architectural decisions around some of the technology choices, to either preserve, in this case, preserve your case, and others, people are redesigning their systems around these AI factories. Maybe they're more enterprise-like in those use cases. But there's two things going on, I hear a lot of acceleration, we've all seen, let's have an accelerator going in, and then offload, you mentioned offload. So how do you guys see those playing in? Because AI's got an acceleration aspect to it, and also the offload benefits, where data could fly off and other systems can connect in without compromising either SLAs or latency requirements. How do you see acceleration and offload going on around your system?

Tim Carmody

>> Yeah. It also, I think, depends on where you are in a particular trade lifecycle. So prior to a trade being able to pull AI in from other sources and being able to make the trader more efficient, really augment, give them everything they need upfront to make them do better trades, and I can give some examples in that in a minute, is offloading from other systems. During the trade, during the actual calls, then it's the opposite, now you're spitting out data to the AI systems that are pulling things in and everything else. And then, after the trade, it becomes about compliance and analysis and everything else, and then it becomes feed the engine again.

>> Yeah. And all those things that check the box, like undifferentiated heavy lifting or boring manual grunt work or toil, whatever, people use different words. I love that toil word, it comes from Google, people use the toil word. I just call it grunt work, basically mundane tasks, like compliance, that's like reports. Okay, so AI's helping you guys. Does the word trader productivity get kicked around a lot?

Tim Carmody

>> Very much so.

>> What does that look like? How do you guys think about trader productivity? Because it rhymes with all the developer conversations, developer productivity, CI/CD pipeline, so they're in the trade channel or trade zone, whatever, they're in trade, they're in process. So productivity, you mentioned dashboards, what are some of the cool things coming, from a either holistic view or deep dive, for productivity and efficiency?

Tim Carmody

>> Yeah. Trader productivity and workflow integration are the two big areas that we're focused on now. That's why we've done a partnership with ICE to enable something called ICE Voice. And you've got to look at it in terms of the different asset classes. So in a lot of asset classes, trades will actually start out on a chat session, and in things like commodities, it's ICE Chat, they're the dominant player in that space. But then they have a toggle tax where they move to another system, an IPC system in some cases, to make the trade actually happen or to elevate it to another point. Well, moving between those systems is a waste of time, and there's also a waste of efficiency for them. And then worse, it becomes harder to do compliance, because now you have broken systems to do the compliance against. So with the partnership with ICE, we've enabled ICE Voice so that people that are on the ICE Chat can automatically elevate it to a call-

>> Seamlessly....

Tim Carmody

>> seamlessly on our systems, and then they can take the transcription of that and put it back into the ICE Chat, and we give them a holistic compliance record-

>> So you basically do the integration for them?

Tim Carmody

>> Exactly.

>> It's seamless to them, they don't miss a beat.

Tim Carmody

>> Exactly.

>> They move from chat to system, even though it's fragmented, not fragmented, disjointed, but you connect it.

Tim Carmody

>> Exactly. And the best part of it, as far as I'm concerned, is that ICE is the leader in their space for the chat part of it, we're the leader in our space for the voice part of it, these are the two industry leaders coming together to deliver a service.

>> Talk about the ICE relationship, because I think this is where we're starting to see a lot of this AI, and robotics is our theme this week, but also other systems. People are connecting, and there's opportunities to have best of product, but also integrate and increase the business model. What's the relationship with ICE about?

Tim Carmody

>> Yeah. So it is just about that, one plus one equals more than two, and being able to bring out more functionality that improves the user experience and the trader productivity and workflow integration and all of that. And it fits nicely into our structure, because we very much believe in an open platform and an API-first philosophy, where integrations and new technologies can all be brought together, and ICE has been a really great partner in that.

>> They have a great technical team too, I got to know a lot of the folks there, so shout out to the International Confidential Exchange, ICE, team. I've got to ask you, what's the coolest thing you're working on right now? Lots happening. You've been living in a steady state world that's been constantly pushing the needle in the market, but there's now this step function change technically. So what's on your mind? What are you thinking about now relative to the business? What's cool that you've got going on? What's the cool things you're chasing down?

Tim Carmody

>> Yeah, sure. So first off, I want to just establish that we're a FinTech company, but more uniquely, we're a TradeTech company. So some of the areas that have been adopted in FinTech, like cloud and other services, are slow to get to TradeTech, because of the sensitivity, because of the security requirements, mission critical. So our customers are now adopting cloud in big ways and we're moving them to the cloud, we have about 15,000 participants already on our cloud product. And just yesterday, we actually end-of-saled our on-premise traditional system and moving everybody to the cloud. So that's definitely a big area for us, which I know might sound a little passe at this point in the rest of the industry.

>> Oh, no, it's great. Really, computing is what it's all about right now.

Tim Carmody

>> Yeah. The other thing we're doing is dramatically changing how people engage with us for other services and being able to put the framework out there that pulls all these things together, including AI-powered applications. So we talked about the head trader app already. The other one that we're coming out with is a AI-curated call list that takes all the information from what's happening in the world, market data, news, geopolitical situations, and then recommends to the trader who they should call, why they should call them, and then during the call, it keeps a list of the sentiment of it and then the market, the diarization of what happens, follow-ups.

>> It truly is a co-pilot model.

Tim Carmody

>> It's a co-pilot model for a trader, and particularly for voice-traded asset classes.

>> All right. Well, great to have you on theCUBE. Thanks for coming in, appreciate it.

Tim Carmody

>> Thank you, John.

>> All right. Tim is in theCUBE here, we are breaking down all the FinTech, all the action. There's so much acceleration, and also new architectures around facility integration, abstracting away complexities. As the world gets faster, the pace of play continues, of course, theCUBE keeps us streaming here, from the NYSE and the NYSE Wired community. With theCUBE, I'm John Furrier, your host of theCUBE, thanks for watching.