Jack Sharry: Hello everyone and welcome. Thanks for joining us for this week's edition of WealthTech on Deck. As you've heard me say many times, I have the privilege of speaking with the best and brightest people in our industry every single day. Today we're gonna talk with someone who passes that test well, if not better than most. And I'm speaking to my friend, John Sweeney, who is our guest today. Learn more about John for those who may not know him. I don't know how they wouldn't know him. I will be talking with John to ask him to share his story around his career evolution, which has been interesting. He's really had a remarkable career. And what is most interesting about the career is how he went from the senior ranks of Fidelity to embark on pursuing startup efforts as he developed concepts and ideas around alternatives, crypto, artificial intelligence. So we'll get into all that in a little bit. So John, we'll fill you in on.
What I love about what you will hear is his willingness to pursue the new, the disruptive, the innovative, the challenging, on the spirit of achieving greater efficiencies and better outcomes for all. John, welcome to WealthTech on Deck.
John Sweeney: Jack, thank you so much for having me.
Jack Sharry: Yeah, my pleasure. Our pleasure. So John, let's start with your career journey. In fact, were off the recording. You were filling me in on a piece I did not know. So please walk us through how you got your start, the trajectory around where you wound up, which led to the product platform head at Fidelity. And then we'll get into what you're doing now in a little bit. But why don't you just talk about that trajectory, just give people a sense of what you've been up to.
John Sweeney: Thanks so much, Jack. You know, began my career actually in marketing. I looked at my clients and saw that they were focused on a lot of things. My client happened to be a large airline and my clients were buying and leasing airplanes and building out new routes. And I said, I want to get to that side of the desk. Went back to business school, looked at what I would call my unique selling point. It was the combination of financial services and the ability to communicate and package. And so ended up at American Express, spent about five years there, working in a series of businesses, which was a fantastic proving ground. Ultimately, my wife was pursuing a doctorate in Boston. And so the decision was we were going to move north. And I interviewed at a couple of different places. I'll say a large traditional asset manager manages institutional money. And the woman I was interviewing with said, what do you like to do? I said, I like to build new things. And she said, then you should go work at Fidelity. So I did.
Jack Sharry: Cool, cool, cool.
John Sweeney: I worked there for about 18 years. Had a fantastic career all across the investment spectrum of businesses. Started at mutual fund product management with the equity funds, moved into the other asset classes. Ultimately then, what we call funds network, distribution of the third party funds through the Fidelity system. From there, it's taking those funds and putting them into client portfolios. We ran the managed account business and grew that. That was a fully integrated business, everything from distribution, marketing, operations, product than planning and guidance tools, trust bank, retirement income. And what I would characterize, Jack, is there was a thread, obviously, of a common DNA of asset allocation and fund and portfolio selection through those, but also a significant layer of technology. A lot of innovation built out. Fidelity's first ETF, first UMA, first SMA. So really enjoyed it. And that innovation and building new things was always sort of front and foremost in my career development choices.
Jack Sharry: Yes. Well, I knew you for much of that time, at least the latter part of your fidelity years. And that was where I coined the term the confluence of digital and human advice in your career as a testament to that. Because all the stuff you just described could not be done without technology. So let's talk a little bit about what you've been doing since, because that's so much about technology, kind of maybe almost in reverse, where the product is important, but you can't do it without the technology. So fill me in a little bit on what you've been up to over the past many years.
John Sweeney: Absolutely. So what I saw about five years ago was the opportunity to put assets on blockchain. And so when I left Fidelity, I joined a firm out in the West Coast, actually moved out there, took an apartment out there, family stayed back East. But it was a firm that was putting home equity lines of credit on an open architecture, financial services, business to business blockchain. And so the really interesting thing was you could put the borrower and all their personal information, credit history, salary earnings, things like that on a blockchain, those would be visible to ultimately the buyer of the loan. The performance of the loan itself. So every Monday at noon on the first of the month, I'm going to debit your bank account and remit that payment immediately to the loan owner. So all the performance is visible to any of the investors. And then ultimately an exchange. And so when loans come off the pipeline every day through the production cycle, institutional buyers could go in and they could actually dial in with very specific granularity some parameters around those loans. Let's say they wanted to get exposure to middle America because they had too much concentration on the coasts, or they looked at somebody who had a high salary but not a long US credit history, so therefore their loan rate might have been a little higher than their peers and their income band. That might have been a good risk for a portfolio manager to try and take saying, think that person's actually going to be able to pay back the loan. So you can dial in with specific granularity through the exchange and not have to have an analyst go through and double check all the numbers because blockchain, remember, is a source of truth. And some of the nodes on our blockchain were credit bureaus, valuation engines on the value of the home. Obviously, the performance is transparent. So it really, in my mind, transforms a lot of the issuance and trading capabilities that I think we're going to see roll through lots of other parts of financial services over the next several years.
Jack Sharry: John, let's talk about your current work. It seems no podcast would be complete without a conversation around AI. Clearly, firms are grappling with what to do with AI and I take advantage of it. A big area of focus for a right turn. I'm sure you do the same.
So let's start at the beginning. How should companies, particularly smaller companies, start on their AI journey? Let's take that through to run its course.
John Sweeney: It's great question. So I'll give you the macro backdrop of AI. And obviously we've seen significant adoption of AI by individuals, but certainly the companies that they work for. And the reason is that AI is what I call visible. To contrast that with blockchain, blockchain is behind the scenes. I don't really see or interact with blockchain per se. AI, I can go on and I can use ChatGPT, I can use Copilot. It's either an interface, for example, almost like a browser that I use.
Jack Sharry: Mm-hmm.
John Sweeney: Or it's embedded in the software that I use. For example, it can read through my calendar or my emails. So as a consumer, I began to understand how it can be applied to a business situation. And what I saw was that clients I was serving were shifting their budgets towards AI because the payback was fairly quick. I mean, almost within 12 to 18 months. The cost of AI software development has come way down and the payback is pretty fast. And so therefore lots of applicability within organizations to use it.
Jack Sharry: Another area that you and I touched on offline, but I'm curious, you just mentioned it, just how things come together between blockchain and AI. I wanted to kind of take us through, think blockchain can be characterized as more of an institutional capability, and that will change as custody, trading, lending platforms of traditional financial firms are played out. AI is more visible, you mentioned that a moment ago, but.
Talk about that. What is the role AI will play if blockchain is background and AI is forefront? What does that look like? How does that work? And if I'm running a business or some part of a business, what do I need to know? What do I need to be aware of?
John Sweeney: Yeah. So I think ultimately, Jack, these two technologies are used in concert with each other. So one of the challenges with AI is that people are concerned about what we call hallucinations. There's some classic ones where some attorneys used AI to write their closing arguments of their case. And AI went out and generated some cases that were fictitious. So that's sort of the most egregious scenario. What I think that we are concerned about in the financial services world is we rely on data purity, efficacy, cetera. So real data is really important. Obviously, if you look at some of the data providers to the financial services industry, they're some of the most highly valued companies in our ecosystem. So what we call verifiable or true data is really important. The great thing about blockchain is that it is a transparent ledger and database that's visible to all the participants. And so when you think at the furthest extreme, the most open architecture, blockchains, Ethereum, et cetera,
We can go on, we can see transactions, we can look at wallets, we can understand holdings. Where financial services firms, particularly those in the US who have more restrictive legislative oversight, they're using perhaps private blockchains or private but permissioned networks where a couple of firms can be on the same blockchain and trade amongst each other. And they can share information packets so you've got visibility. And I think there's some balance of centralization and decentralization that's going to ultimately occur. The bottom line is blockchain becomes this source of truth that's visible to the participants, visible to the regulators. And on top of that, we're going to be able to use artificial intelligence models to look at that data, to analyze it, to make decisions and know that they're working off of a source of truth. It's not some fabricated data that's coming from off network.
Jack Sharry: Talk a little bit about the business that you're currently heading up around AI and I'm assuming blockchain is all part of that. So I you to just describe the business, talk about what you're doing, who you're doing it for, what is all that, how's that all that come together?
John Sweeney: Sure. So the firm is called Praxis Solutions. Praxis is a term that Aristotle coined a couple thousand years ago. But essentially, he was trying to admonish his philosopher friends to get off the steps of the temple and go down into the people and do something with the things that they're writing about. So our tagline is think, make, do. And thinking is really strategy. Half our team are business practitioners like myself who've run businesses and can walk into a client and help them and if all we did was write a document and leave it on their desk, that would be helpful to some of them. The real value add though are the other half of the team who are the AI scientists led by a CTO who spent a large part of his career as head of a brokerage business and brokerage technology firm, as well as a chief data officer within the firm. So knows data, knows brokerage technology. When he looks at his peers in the CTO seats at our clients, it's a peer to peer conversation.
It's not an AI theorist saying, don't understand what an RIA is or a basis point is. He's done all these operations himself, but also is an AI professor and has a team of PhDs in AI who can help execute these sets of capabilities. And the third part, most importantly, Jack could say is what we call the do part. And so a lot of times software, SaaS software, they'll leave a piece of software on a client's desk, but it's not implemented. It's not integrated.
There are all kinds of examples of even the largest SaaS firms in the industries having consulting firms who basically draft behind them and implement and integrate SaaS software into their ecosystems, customize it for their environments. And so what we do is actually build custom software for an operating environment. We work within the existing bits of software that clients have, A, connect them, B, drive workflows, but C, what we're trying to do is not build functionality that's packaging or is what I'll call the arms race of features where every year I have to come out with a new release and add a few new bells and whistles that probably aren't that important to most clients. They want to replicate existing workflows, improve them, and make sure that they move more seamlessly. And so ultimately what this does is allows us to integrate within their operating system and move data and decisioning more quickly and move from data to information, ultimately to knowledge.
John Sweeney: And so you think about knowledge transfer through an organization as organizations change, as people leave, as they change jobs. You want to make sure that that knowledge is resident within the organization. It now can be using AI.
Jack Sharry: So it sounds to me like AI is part of a problem-solving discussion, orientation, exercise. And certainly don't have to name names, would want to. But talk a little bit about the types of firms you work with, the types of issues, what are the pain points that you're addressing? So take us through the, I get the big picture of it now. So what do you do about it? What does that do part? do you, give me some examples of the sorts of things that you help them with.
John Sweeney: Great question. actually, Jack, it asked the question, where do we start? Right? So how do we begin to engage with your firm? And because we are a consulting firm, we can help you with some things that are as simple as business process re-engineering or an AI strategy. And so for some firms that are early in their journey on AI, they want help in those specific areas. For firms that are further along and they've perhaps used some of what I call the office productivity capabilities.
Look through my emails and figure out all the the ones that I need to respond to and draft a response That's a fairly simple what I'll call office productivity type of exercise We don't build that capability because we think that the office productivity software firms will do that Well, we focus on the types of work that are germane to financial services firms I'll say the first two categories we work with wealth managers and asset managers. They're two sides of the same coin and you have asset managers certainly trying to reach wealth managers and educate them on their products. So there's a natural ecosystem between those two. And what I find is that we have capabilities that often sit between those two distribution organizations. I'll give you an example. We had a client who sells private real estate deals. They would see about 20 new deals a week. They had their human analysts reviewing them and it took 10 days for each deal to get through a review system where they could extract relevant information from 1,000 pages of legal documents and spreadsheets and presentations, summarize it and put it into an output that met the investment committee's criteria and was consistent with what their compliance organization wanted to see. So we worked backwards from that desired output and said, what information do you need us to extract and how do we A, summarize it and then B, generate an investment memo? And we got that down to actually 10 minutes.
It still takes three days to get through the operational systems because there are humans in the loop reviewing the decisions and making decisions on their own, which we think is appropriate. But think about that capability of analytics. It involves a large language model, involves extracting data from charts, graphs, spreadsheets, et cetera. It requires summarization and production of new material. And so there are a couple of different AI capabilities that are resident in that piece of work.
John Sweeney: In this particular case, it was a broker dealer looking for approval. Is this basically a suitable investment for my clients to offer or my representatives to offer to clients? That's one level of criteria. The same engine jack can be turned for an asset manager who's reviewing real estate deals or private equity, private credit deals, and they can be looking for the best investment. And so it's just tuning the output to be customized to the output requirements of the client. But the engine remains quite similar.
I need to extract the same types of information. I just need to weight it differently for an asset manager making a binary investor don't decision versus a suitability decision to say all these above this threshold are suitable.
Jack Sharry: Gotcha. So if I'm hearing you, people like yourself that are helping to design the solving of the problem or solving of the issue or addressing the pain point, and then your data scientists then do the software development to make that happen and to fill me in on all that.
John Sweeney: Sure. So what's really interesting to think about AI as a development tool as well. So we think of it obviously as a front end and we draw a pyramid for clients and at the lowest level are the chips, right? So Nvidia has obviously been in the news, all the chip manufacturers building very fast functional chips. The next layer up, the engines that we're all familiar with, the OpenAI, the co-pilots, et cetera. What you start to get above that are layers of infrastructure and then apps.
And I would use the same analogy with the HTTP protocols built way back in the 60s by DARPA. And when they ultimately got commercialized in probably the mid 90s through Netscape and browsers, those are really application layers on top of it. And what you and I are experiencing today over this video chat is yet another application built on top of a network communication protocol. is very similar in terms of its structure. And we're playing at the application layer where our clients are using software that we're building on those ecosystems to help deliver functions and business processes that improve their economics and make their client experiences better and make their operations run more efficiently. If I may, yeah, go ahead.
Jack Sharry: So, help me understand what's the sort of business model? Do you come in as a consultant? you come in? Frankly, how do you get paid to do what you do? And I'm trying to button it down to what does the work program look like?
John Sweeney: Yep. So some clients say, Hey, I need help with consulting and it can be around data architecture, data governance. can be all sorts of things where they just need help. And we've got folks in house who can help map data, structure it and organize access to it and basically be able to just, you know, clean things up. That's one piece of work. Sometimes the strategy work encompasses product design and development. We have an asset management client who is non a U S based.
John Sweeney: They wanted to come to the United States, didn't know the networks, didn't know the access points, really had to look at the competitive environment. So that was more product consulting. The sweet spot is really the software development. And so again, we key around a couple of different we call agents that replicate work within different operational functions. And so I mentioned the analytical one.
We have ones that are built that do marketing. We have ones that are built that do compliance. We have ones that analyze equities, things like that. Where this ultimately goes, Jack, is ongoing training. And much like a human analyst that you might hire out of school who took accounting classes and knows finance and understands Excel, they don't know your firm's investment philosophy. And they don't know how your firm makes investment decisions or what's important to your fund and how to screen for those criteria. So what the engine does is it gets trained and you should think of it the same way you think about a human analyst that you hire. They're not going to be the level of a vice president on their third day of work. You have to train them, teach them, give them feedback along the way so that the business model is essentially three tiers. One is consulting, two is the software development, three is the training and ongoing integration of the software over time.
Jack Sharry: And how do you see this playing out? Or how is it playing out? it that, I would assume you start more on a consulting basis, try to scope out the problem, the challenge, the issue, whatever it is that they're grappling with, the immediate pain point, it sounds like a starting point. And then over time, does the relationship evolve so that you are doing ongoing consulting in other areas? Talk a little bit about what that journey looks like.
John Sweeney: Depends. so what you have are you have people at different points of, I'll say, operational efficiency and growth. Jack, I haven't met a client yet that didn't have a growth opportunity. so I'll say almost all of our conversations inevitably gravitate to the how do you help me grow my business and the equation.
Jack Sharry: Let's say check they don't know how to do that part
John Sweeney: The way we try and frame the solution sets are what I would characterize as front, middle, and back office. And so front, I characterize everything to do with the client. That's essentially the growth engine. How do I target market position, communicate? We actually bought a call center. So think of it like an inside sales desk. So if you're an asset manager, we can actually execute back to that think, make, do framework.
Jack Sharry: Right.
John Sweeney: We can do for you. We can supplement your inside sales team and help them operate more efficiently by getting them physical appointments with intermediaries who are interested in their products. So think of that full stack of capabilities as the growth oriented part of the equation. The middle column, what I would call is sort of the investment secret sauce, scaling that process, but making it customized and making it feel like we're not coming in with our model telling you we can use AI to beat your benchmark.
Jack Sharry: Yep, yep.
John Sweeney: It's making your process more scalable. And a great example there would be a small emerging market manager who with a couple of analysts is tracking stocks in 130 countries around the world. Think about it only in US or English language, but the real data probably resides somewhere in local language. So can we use AI to A, translate it, B, normalize it to English language across multiple different sources?
And then C, bubble up the information to the portfolio managers and analysts so they can track and make decisions on the investments they're following. We're not making the investment decision for them. We are basically surfacing the information so that they can scale their operations process. Exactly right. Back end compliance operations IT. Again, we hired a head of compliance and he's building out compliance modules because nothing that we create in the front end on the marketing engine can get out the door unless it goes through
Jack Sharry: Sure, Enhancing their process.
John Sweeney: Through compliance. Jack, where these all string together is the agents independently replicate human work functions, but ultimately they operate like an organization. So they have to try and talk to each other in what we call a multi-agent platform. And that's really where you're stringing together multiple agents who act in concert within the parameters and rules of the organization. So when we create marketing materials and it comes out the other end of the pipe, you know it's already been through the compliance engine because
If it hadn't, it would have iterated back and forth much the way that the marketing department and the compliance team today currently iterate. They walk documents back and forth down the hall. Today that happens dynamically and within the software.
Jack Sharry: Sure. Yep, yep. And so how does the model work? you train the trainer? Do they set up departments and you work with the head of the department or the team or whatever? To talk a little bit about what that dynamic is. How do you enable them to succeed over time?
John Sweeney: Yeah, so that's a great question. Think about legal. Legally, AI in general is great at taking large complex data sets that are today analyzed by expensive humans. And so think about legal, right? So when the SEC comes out with a new law, or FINRA has a new law, you've got a lot of lawyers digesting it, summarizing it, and then more specifically trying to apply it to your firm's policies and procedures. So our compliance module essentially looks at the law, that's one, I'll say, piece of work or one bot that the agent performs. The second piece then would be compare it to my current policies and procedures, understand the deltas between what I'm doing today and what the new rules are. The third stage is write the new policies and procedure for me that conform to the new law. And the fourth stage is build an audit module that goes back and make sure that my operation does adhere to the new law. So you've got a couple of stage process.
That when you work through, begin to understand, these are jobs today that people are doing, and I'm working within an organization's framework to understand the custom and specific policies and procedures that they have, and making sure that they're compliant with a consistent set of laws.
Jack Sharry: How does that work on an ongoing basis? This, by the way, very fascinating. Everyone talks about AI. I mostly don't know what they're talking about because it's too abstract and too hard to get your head around. But from what I'm hearing is you can set all this up. How does the ongoing management, what does that look like? Are you guys part of that? Are they part of that? you enable them? What does that look
John Sweeney: Yeah. So that is part of that, what I'll call training process. And so you should think of that as much like the ongoing training classes that we all take for licensing exams. Rules change, situations change, obviously market data changes, business prioritizations change. So you want to continuously train the model to make sure that it's current with the current environment and business practices and laws. And so that ongoing training process is what we offer to clients.
And that's what I call the third leg of the revenue stool from our standpoint. We want to make sure that we're delivering something that's complete and packaged for clients, but we feel that there's a value added service to ensure that it's continuously updated. Then the organization owns that knowledge. It doesn't reside in the individuals who punch in and punch out every day.
Jack Sharry: Gotcha, gotcha, interesting. So we've covered a lot of ground, two questions left to go before we say farewell for now, maybe more, we'll see. Where do you see all this going? Obviously you're busy in the, it sounds to me like startup, early stage, getting people up and running. Where do you see it going? And then we'll talk about some other things in a moment, but let's start with that.
John Sweeney: Yeah. So, you know, a couple of things. We've got to deliver for the clients. That's sort of, you know, job one. We've got a great set of clients that have entrusted us with an emerging technology. And so they're asking us to be their guide, their Sherpa through the navigation of new technologies and applying it to their systems. Those have to have incredible integrity. So back to that point of what I call sources of truth, we've got to make sure that we're delivering something that's at least at the outset, at par with what their humans are doing and give their teams the ability to review. Ultimately, we want to be at the place where we're delivering something that's of superior value to the organization. Secondly, think the second part is to take additional capabilities. And as we think about the operations, we're working with clients on what I'll call the second and third and projects within an organization. So we recognize when people see value in what AI software can do for their organization, they say, well, how about this over here? And you begin to build converts and advocates within the organizations rather than what I'll say, reticent buyers who somebody told them to take a look at this. They then see the value of it. And Jack, what I think has to happen is we have to change how we think about what is, you know, I'll say power or within an organization today, its size of budget, its size of team, its scope of responsibilities.
John Sweeney: And there are theories out there where people are saying, how can I run an organization with relatively few people, but run it incredibly efficiently? What that's going to do is that's going to bring then the cost of advice down. It's going to democratize advice. We're going to be able to offer higher levels of service with human advisors to more people at a cost effective price. And so ultimately, when we think about our aspiration as a firm and why we think AI is going to be a useful tool for financial advisors in the ecosystem that serves end investors, we think it democratizes a lot of access to institutional grade theses.
Jack Sharry: Interesting, interesting. So we've covered a lot of ground. What haven't I asked? What's been left out so far? I'm sure there's plenty, but anything in particular you want to make sure our audience is aware of?
John Sweeney: Jack, I'd say that the one thing to think about is that the cost of custom has changed dramatically. So 10 years ago, if you went in to buy a custom suit, they'd measure you, they'd go back and forth a few times with some templates, and then ultimately, six weeks later, maybe you'd get a custom suit back. The cost of custom software used to be very expensive, and AI development has changed that. And so the best analogy I could provide is one that I borrowed from a speaker at another conference.
But he said, 10 years ago, if you and I were playing a game of chess, I would have had to program what the knight did, how the bishop moved, how the queen moved. I would have had to program each of those pieces. Today, I ask the AI software to look up the rules of chess and then program a game. Furthermore, I spent a large part of my technology budget and prior roles doing quality assurance. Now I write an agent to go through the software and make sure that it works effectively as I've designed it.
So the cost of software development has come down tremendously. No longer do we have this expensive J curve of development time where I have to spend a couple of years building a piece of software. I pop out at the other end and I hope that I've got something that's still relevant in the market. I can iterate very quickly and deliver a prototype that the client can react to and back to this training and feedback mechanism that creates a learning organization.
We can do that in a relatively short period of time. And what that means is the payback period is really short for these organizations. So that's a really different shift and a paradigm shift for the buyers of these services to consider when they think about custom and their perception that custom is expensive.
Jack Sharry: Interesting. So in addition to that sounds like a key takeaway. Any key takeaways you want to leave with our audience in terms of the future of AI?
John Sweeney: Yeah, I'd say, know, if you're reticent and not engaged in AI in your organization, we certainly see some organizations who have prevented their employees from engaging in AI with company data for good reasons. Talk to somebody who can help you organize the data. There are certainly ways that you can wall off your proprietary data and not expose it to your competitors. So I think that's really important to consider. The second thing is then work on projects that I would characterize in a two by two as having high impact and relatively easy for the organization to execute. And the third part is I use the term, do you want AI to be visible to your organization or invisible? So if we think about the chat engines that we interface with, that chat GPT interface, copilot interface where I put in a prompt, that's visible AI. I know I'm interacting with AI and it will look and feel different. And my answer will be different than the answer I get from a search engine.
Invisible AI may operate behind the scenes and it will be served up to your users within your corporate ecosystem in a way that may not be visible to the users, but the IT organization may be comfortable with the way the process is delivered and the the information is delivered. The end users may just look at it as a superior client experience.
Jack Sharry: Interesting, interesting. Well, this has been fascinating as we've had many guests on the topic of AI. I haven't got much beyond the platitudes. You've been very specific about what it does, how it works, what it can do. And it also sounds like we're just scratching the surface. So one final question before we bid adieu to one another. And this is always my favorite question on our podcast. What do you do outside of work, John? Something that you're excited or passionate about that people might find to be interesting or surprising.
John Sweeney: I have wonderful family. My wife teaches me a lot. She has a doctorate in systems thinking. And so I learn a lot just from the dinner table. We have three wonderful children, two of whom are in the business, one still in college. And it's quite interesting, Jack, to see them come back with questions, obviously, where dad actually knows something now, but that's wonderful. And then just recreationally, I play hockey twice a week. I live in outside Boston, so play a couple times with a bunch of folks early in the morning. And that's what keeps me active and youthful. So this is the days of the week that I'm happiest.
Jack Sharry: That's great. That's great. Yes, I like it. like it. So John, thanks. It's been a great conversation. I've learned a lot. I really appreciate it for audience. Thank you for tuning in today. If you've enjoyed our podcast, please rate, review, subscribe and share what we're doing here at WealthTech on Deck. We're available wherever you get your podcasts. should also check us out at our dedicated website, wealthtechontech.com. All of our episodes are there along with blogs and curated content for many folks around the industry. John, thanks a lot. This has been a lot of fun. I really enjoyed it.
John Sweeney: Jack, thanks so much for having me. Appreciate it.
