Tim Gasper [00:00:32]:
Hello, everyone. It's time once again for Catalog & Cocktails. It's your honest, no-BS, non salesy conversation about enterprise data management. With tasty beverages in hand, I'm Tim Gasper, longtime data nerd, product guy, customer guy at data.world, joined by Juan Sequeda. Hey, Juan.
Juan Sequeda [00:00:49]:
Tim, how are you doing? It is a pleasure. Wednesday, middle of the week, towards the end of your day. And here we are, another live episode of Catalog & Cocktails to talk about data. And today, super exciting to have Colleen Tartow, who's a field CTO and head of strategy at VAST Data. And what I love about having a guest like Colleen is because I love being at conferences, meeting people, being at their talks, and like heck, we need to continue this conversation. So, Colleen, it is great to have you finally on the podcast. How are you doing?
Colleen Tartow [00:01:19]:
I'm great. I'm super excited to be here. Thanks, guys. Thanks for having me.
Juan Sequeda [00:01:23]:
Yeah.
Tim Gasper [00:01:24]:
So cool to have you here. I remember seeing your talk at Data Council here in Austin around the AI stack and like, how to really get value out of data, and it was so cool. Glad we could chat and so excited we get to continue the conversation here for all our wonderful viewers.
Colleen Tartow [00:01:41]:
I'm thrilled. Yeah.
Juan Sequeda [00:01:42]:
Let's kick it off with our tell and toast. So what are we drinking and what are we toasting for?
Colleen Tartow [00:01:47]:
I am drinking a little bit of Prosecco that I found in the fridge, which a couple days old maybe, but still bubbly. So we're good. And I believe we're toasting to October birthdays. Right? Today's your daughter's birthday. That's your birthday. My birthday is in a couple weeks and.
Juan Sequeda [00:02:04]:
And tomorrow is my other daughter's half birthday, so.
Colleen Tartow [00:02:08]:
Oh, half birthdays count, so definitely. Yeah.
Juan Sequeda [00:02:11]:
October. October. And I'm also toast that it is. We're in Austin and it is the most amazing, beautiful day I have. Just.
Tim Gasper [00:02:19]:
It is a wonderful day today.
Juan Sequeda [00:02:20]:
So it's one of those days you're like, you know what, people? We complain about the heat and everything, but the heat's only like 3 months, 4 months, rest of the year just like this. It's beautiful.
Colleen Tartow [00:02:29]:
Oh, that's nice.
Juan Sequeda [00:02:29]:
Tim, Tim, what are you drinking? What are you toasting for?
Tim Gasper [00:02:32]:
I am drinking actually a mocktail today. I know a little Unusual for myself. It is a recess island spritz. Not too shabby. I do recommend it and I'll cheers to October birthdays. I'll give you all my cheers.
Juan Sequeda [00:02:45]:
Well, by the way, you're doing, I like you're testing all these different mocktails and these are really fascinating. I need to get.
Tim Gasper [00:02:51]:
Yeah, I'm trying them out, you know, because, you know, I drink a lot of cocktails for those who listen to the show. You know, I drink a lot of cocktails, but I want to be able to mix it up and have some mocktails every once in a while. Right. And so I'm trying to figure out which ones I like.
Juan Sequeda [00:03:04]:
Well, I'm on the last bit of the apparel that somebody left in my house. I don't know when, I don't know who, I don't know when. And I'm doing an aperol Cosmo. I'm like, I haven't done vodka yet, so this is a better vodka and aperol. And here we go. So, hey, cheer. Cheers to October babies.
Tim Gasper [00:03:17]:
Cheers to October babies.
Colleen Tartow [00:03:19]:
Cheers.
Juan Sequeda [00:03:19]:
So our topic today is streamlining data value. So our warm up question is when you streamline something, you're simplifying it to make it more efficient. What is something that should be more streamlined in our day to day.
Colleen Tartow [00:03:32]:
I'm probably saying this because it's 5:00 and my kids are going to be asking for dinner as soon as I get out of this. I think dinner should be streamlined and I think that we've been trying to do that with these things like hello Fresh and blue apron and purple carrot and all these like, you know, meal boxes and like, you know, Trader Joe's prepared meals and stuff. And we've been trying to find that like perfect balance between fast and easy and nutritious and tastes good and all that and like my kids leader, that kind of thing. And so that's what we're trying to streamline because I honestly like could make four dinners, one for each of us and that would be what would make everybody happy. But I'm not going to do that. So, yeah, that's something I want to streamline.
Tim Gasper [00:04:16]:
I can definitely feel you there. I, I made some Italian wedding suit this last weekend for the first time and the recipe said it was going to take an hour and 15 minutes and it definitely took me three. So, you know, it's hard to streamline all that, especially if you got different tastes in your household. Right.
Colleen Tartow [00:04:32]:
So I've streamlined soup to the point where I literally just roast a bunch of Veggies. I'm vegetarian. I roast a bunch of veggies, like whatever I get at the farmer's market. It doesn't matter what it is. Just chop them off with the same size olive oil, salt, pepper, roast them all together with like a head of garlic, throw it in a pot with like broth, and then just go at it with the immersion blender. And it's amazing. Every time it's different.
Juan Sequeda [00:04:56]:
Every time I do the same thing. That's an easy way to do it. And then I just keep a bunch of it. And then, you know, it's easy. I just drink out of a cup. I don't want to go play. That's how I'm streamlining drinking.
Colleen Tartow [00:05:06]:
I love that. Look at that, streamlining. I love that.
Juan Sequeda [00:05:09]:
Tim, how about you stream. What do you streamline in your day?
Tim Gasper [00:05:12]:
You know, honestly, a little bit similar to dinner, but just in a different context is syncing up schedules.
Juan Sequeda [00:05:17]:
Oh, my God.
Tim Gasper [00:05:18]:
Trying to schedule and time stuff with people. Like, even if you have an EA or somebody who's helping to coordinate, it's always like, well, what about this? What about that? Oh, an hour. Oh, wait, no, somebody can't make it. Oh, okay, let's reschedule. And like, oh, man, I wish it was simpler.
Juan Sequeda [00:05:30]:
That was. And then we have these things like, what is it? Clockwise and blah. And then you have all these doodles and stuff. But that stuff is never.
Tim Gasper [00:05:36]:
Yeah, it helps a little, but also, you know, like pros and cons, right?
Colleen Tartow [00:05:39]:
Yeah, but no, we have personal calendars and our work calendars and the kids calendar and the family. And it's like, yeah, how do you reconcile? Figure that out and let me know.
Juan Sequeda [00:05:50]:
All right, well, let's talk about others type of streamlining. So, Colleen, honest ops, what is the fastest way you're seeing to streamline the path to data value?
Colleen Tartow [00:06:00]:
People, I think, are starting to consolidate things. There's this modern data stack madness that we've had the past, like, however many years, and there's literally a tool for everything. I mean, everybody's seen the Matt Turkish MAD data diagram, right? And you. There's thousands of tools on there. And it's like, obviously you don't use all of them, but like, things have gotten composable to a ridiculous degree. And I'm seeing folks, especially enterprises, start to consolidate or start to. Or they're stopping the drive towards that composability, right? And like, you still want some choice and you want some optionality and you want composability and you want future proofing. Et cetera, et cetera. But like there are good vendors out there who have solid stacks that are less complex, right? And so streamlining is all about simplification and it's all about really focusing on the value. And so I think that's, that's what I'm seeing anyway, and that's what I hope we do because I think it makes a lot of sense.
Juan Sequeda [00:07:04]:
So it's interesting that you, you started here with the kind of on the tech side, which I would agree that, that because we focus so much on tech, right? And then the tech becomes like, we partition all these things and then we just focus so much on the tech side and then that's why we're not helping it. So consolidating the tech stack is one thing, but what's coming afterwards that.
Colleen Tartow [00:07:25]:
Well, I think the idea is that the tech has been almost a blocker towards success for getting value added data, right? Like you have this pipeline of data like you, from start to end, you know, you're putting it through all these paces so that you can curate it to make it make sense in terms of the business. And so by consolidating the tech stack, you can actually, you know, shorten that path. But then you can also really focus on what's specific to your business, which is the specific curation, right? And that's really the name of the game is like focusing on that, you know, whether it's data warehouse or data Lake or whatever you're doing, or Lake House, you know, focusing on that curation of data into the value, right? And making it consolidated, et cetera. But then there's also like this addition of unstructured data and AI and all that. And so as that's coming in, it's making something, it's making things even more complex. It's like, do you have two pipelines? Do you try to do everything in the old stack and the new stack? And so I think what people need to do is really focus on that end value, right? Like what is the question you're trying to answer and then work back from there.
Tim Gasper [00:08:36]:
Yeah, focus on the end value is obviously really important. You know, is that getting muddier now? As you know, we kind of enter this AI age. You mentioned, like unstructured documents and things like that that we're trying to feed into AI and you know, is that a data problem? Is that an application problem? I don't know. Like, I think that's a lot that we're trying to reconcile there. And there's a lot of just Technical excitement about like, oh, we got to get our AI strategy moving forward. Well, you know, I certainly hope that most companies are writing their AI strategy like a use case oriented document and not like a, well, we got to make sure we set up a vector database. Right. But I'm curious, like, what's your take, Colleen, on like, what's going well and what's going poorly as it comes to like driving business value when it comes to building out your stack?
Colleen Tartow [00:09:22]:
I think so. What's interesting to me is that with AI, a lot of people are like, everybody goes to the chat bot as the first example, right? Like, because they're like, oh, we already have chat GPT.
Tim Gasper [00:09:32]:
We can just, you know, just make it, learn our business and then it'll be our little friend. Right?
Colleen Tartow [00:09:39]:
Fire all our support people. It's a great idea. But I think the challenge is that a lot of folks, especially at the exact level are like, we need to do AI. AI is the future. Which I don't disagree with. We need to do AI. They're not focusing on how they get value out of the AI, right? Like, you don't want to just do it for the sake of doing it. And you, you don't want to go buy a GPU and not know what to do with it. You don't want to rent a GPU in a server farm somewhere and not know what to do with it. I think the challenge is that you want to really understand that use case like you said. And then also once you have the use cases, you want to leave room for exploration and experimentation and all that good stuff. But you can't start there the way you could at smaller volumes because you're talking about such large resources here that you have to think about. I also think there's just like this sort of dearth of creativity right now. And it's frustrating because it's like creativity is so much of what's interesting about AI, right? Like the thing, like, I don't even think we're scratching the surface of what we can do with AI.
Tim Gasper [00:10:49]:
What's an example of something you would say is, you know, not, not really thinking creatively, maybe thinking kind of the boring or underwhelming versus, versus what's a creative. Wait, what would you say is like a creative use case?
Colleen Tartow [00:11:02]:
I mean, I don't want to knock anyone because I'm sure trying their best.
Juan Sequeda [00:11:06]:
But yes, honest no-BS
Colleen Tartow [00:11:08]:
I mean, honest no-BS. I think all the like natural language to SQL stuff, I'm like, dude, I like SQL. It's easy for me to write, I don't want to have to debug ChatGPT SQL, right? And the chatbot's another example. But it's like, you know, there's some cases where it's okay for me to write SQL and I don't need natural language to SQL. And I think everybody's shoving that in to be like, look, we're doing AI. It's like that was unnecessary. Right? There's not a ton of value in that because SQL was never the blocker to success. Right? Like, SQL is not the problem. The problem is that, you know, the rest of the pipeline is challenging and like we don't have the data we thought we had or we have all this data and we don't have the resources or you know, it's people problems, it's things like that. But you know, there's definitely places where, you know, ChatGPT or any AI gen AI is going to be really useful. So I just think that we're some, some organizations are getting more mature. Like I think a lot of the pharma companies are really pushing the boundaries of what you can do, which is amazing, you know, with like customized drugs and things like that. But I, I think there's the other side of it where it's like, you know, a board member says, are we doing AI? And that trickles down to like a bunch of engineers trying to do AI in a product that doesn't really need it.
Juan Sequeda [00:12:29]:
Let's expand more on, on this because I, I, I find it fascinating. You said there's the, with AI right now we've gone into like this lack of creativity. We're kind of like doing the, the obvious things, which in a way makes sense. Right. It's kind of the, it's the obvious quote unquote, low hanging fruit type of stuff, which is, we realize it's not that low hanging fruit, which is like what are other things? So I mean, you mentioned pharma. I've said like you, you talk to a lot of, a lot of folks in the industry, like what are the stuff that you're seeing and are actually, let's, let's challenge the listeners on like what should they be thinking about? Let's just, let's be creative here.
Colleen Tartow [00:13:04]:
Yeah, I mean, well, that's the thing is it's like hard to be, it's hard to force creativity. It's hard to be like, hey, be creative. Right? Like inspiration strikes where it strikes. And I think innovation comes from having like a diverse workforce. It comes from all These different places. And so, you know, organizations that have that and have, you know, fostered innovation and creativity will do better at things like this. That said, you know, I think that's why Apple, you know, Tesla or whoever, you know, they've got a lot of people who are focusing on that and pushing the boundaries of that in a way that an insurance company might not be right, like an older legacy company with legacy hardware and legacy thought processes, if you will, you know, But I think there are a lot of really interesting startups incubating right now that are going to blow our minds in like five years, maybe, maybe even faster, maybe two years. So.
Juan Sequeda [00:14:01]:
Yeah. Are you, are you. So I'm hearing this, I'm thinking like what type of organizations actually have the, call it the, the right basically to say you, you deserve to go spend time to be creative versus like you actually don't, like, don't do that. You need to go focus more on like very specific things right now. Like that's probably distraction, right? Or, or am I pushing too much here?
Colleen Tartow [00:14:25]:
Well, I don't know. I mean I, I would like to think that anybody has the capacity for that creativity. Right? But that said, like, not everyone's gonna have the tools in the background to understand what's possible. So I think a lot of people are learning about AI right now. I think there's like a huge push for people to like understand what rag is right, like understand what agents are, understand what's possible. Because we all play with ChatGPT, so we all understand it at some level or perplexity or whichever your favorite GPT type thing is. But I do think that there's the everyday, like, oh, I use this to craft an email to someone versus how can this make our product better? And so you have to be very familiar with it to understand what's possible and what's feasible for a product. So I think we're sort of at that phase right now. And then there's all the backing behind it, like the infrastructure and building out, you know, making sure your data is there and what's possible, which in a lot of ways is the same problem we had with data science 10 years ago. Right.
Tim Gasper [00:15:29]:
Are companies struggling on both the technology front as well as the use case front right now when it comes to things like AI?
Colleen Tartow [00:15:36]:
I think so, I think so. Like, I think there are companies that have great ideas for both. Right? Like, I don't think that's rare. And I think, you know, there's definitely like percentages floating around. Like 70% of companies are doing X well, 40% are actually successful or whatever. And you know, I see these numbers float by and I'm like, I don't know how you're measuring that because like the successful with AI mean you threw in a chat bot. I don't know. But I do think that everyone's trying to do better at both the data side of it and the infrastructure side of it and the creativity side of it. Right. And it. There's something that needs to, there's a muscle that needs to build between me, let's do AI and the actual implementation.
Tim Gasper [00:16:22]:
Yeah, that makes sense. We have an interesting question here coming from LinkedIn actually where Bill says, what are your thoughts of AI and the role that it plays in decision support? I think this ties to some of the use case conversation of what are some of the use cases or the mediums or how you work it into your workflow that you're seeing be effective around AI.
Colleen Tartow [00:16:47]:
Yeah, I think it's funny because, you know, I was talking to my husband about how like our kids are going to have the ability to use AI when they write essays in school and how are we going to police that and stuff. And it's funny because I was saying like we didn't have that, so we learned that sort of critical thinking skill of questioning. Does this answer from perplexity make sense? Right. And so I think that we still need those critical thinking skills and we still need the human in the loop to sort of judge whether or not this answer makes sense or not. And so I think that you can have decision support but you can't have it be completely autonomous.
Tim Gasper [00:17:32]:
Yeah.
Colleen Tartow [00:17:32]:
You know, like I don't, I don't trust the Internet and the Internet is what drives all of these gen AI models. Right.
Juan Sequeda [00:17:41]:
So like, but this is important to realize and acknowledge that this isn't just something that comes up because of AI today. I mean there was a reason why humanity and people didn't want books to be released to the world. Right. When the press came out they said, well, people can write things and you don't know what they're going to go, right. Then they're going to spread things like, like this happens over and over again. I remember once somebody like people were against the typewriter because they're going to say I don't know who wrote that because before people were right with the pen or whatever and they're like, oh, I know who actually wrote this now. Anybody. Like, so this is constantly showed up in humanity over and over and over again. So this isn't new and people like. And it's like, oh, you can't believe this stuff. Because, I mean, this has always been the case. I mean, we've been using the web for the last 20 plus years, all right? So I think we sometimes kind of forget that, that get out of our own bubble in a way and also see what we've done before. And, I mean, I have faith in humanity, so.
Colleen Tartow [00:18:41]:
We're ultimately good. I will say, though, that I think the piece that's scary and unique this time around is that there is a lack of transparency. It's very much a black box. Like somebody writes a book. They wrote a book, they took words and put them on paper. Right. Whereas most people. There's like 40 people in the world who can write an LLM, right. And, you know, maybe more, but they're all at Stanford or whatever. And so, like, you know, you have these people who understand at a very deep level how it works. And most people don't. I certainly don't. Right. And so I think that there is a very high barrier to entry, to being an expert in AI in a way that previous technologies maybe didn't have. And so I think that's a challenge that. You know, I talk to my parents, like, they don't understand this. My dad's an economist, Right. Like, he doesn't. He understands math, but he doesn't understand this. And so it's just so far beyond the understanding of a lot of people that they're. They're nervous about it, which is good. They should be. But that doesn't mean you should just write it off either. It means that, you know, we need to start educating people. We need to start learning more about how to bring it to the masses and how to regulate it, et cetera.
Tim Gasper [00:19:57]:
Yeah, No, I think that's a really interesting point. It makes me wonder a little bit, right? Like, supposedly there are hundreds of millions of users of ChatGPT and it's like the fastest growing service that ever, you know, lived in, you know, ever came into existence. Right. In terms of user growth. But then to reconcile that, I agree, like, there are tons of people I talk to who, like, don't really get it. Like, they're just starting to tap into whether they're more technical or not. Right. Or they're more. Or they're less technical. And it makes me wonder a little bit, like, what are they doing with Genai? Are they. It makes me think about. I wildly underutilize my Alexa. My Alexa is basically a kitchen timer. For me. And so are lots of people just using genai, like a kitchen timer. Is that kind of what's happening right now?
Colleen Tartow [00:20:45]:
I think so. I think a lot of people haven't really thought about the capabilities of it. Like, I was writing a talk last week and I probably shouldn't have put this down on air. I was writing a talk last week and I kind of had an outline in mind and I went to Gemini and I was like, can you create me a deck that'll do, that'll talk through this? And it was like, sure, Here's a potential outline. What do you think? Like, let's go through the outline together and then I'll create the deck. And in the end, I ended up not using it, but I use the outline right. And it's like having a conversation with someone who understands you.
Juan Sequeda [00:21:16]:
Yeah.
Colleen Tartow [00:21:17]:
That said, I tried some, like, manual rag to try to get ChatGPT to like, write like me. I gave it a lot of my writing and was like, here's some examples of my writing style. Can you do this? And it could not get my writing style.
Tim Gasper [00:21:30]:
Yeah, I think it can be pretty difficult. And I know a lot of companies, including Data World, like, are really trying to figure out how to get this rag thing right. And it's hard. It's really hard. And where it seems like, you know, to go back to Bill's question here where, like, the decision support thing is really effective now, it's not too. It's a little bit harder to get it to interact with your data and give you these faithful, explainable, break the black box kind of answers. But what it is really good at is helping you to clarify your thoughts. Like having a little bit of a debate partner. I remember just the other day I was working on some product prioritization for our company and actually passed some kind of information about, hey, here's some of the different product projects that we're thinking about. How would you stack rank it if you were going to stack rank it? Right? And it said, well, here's how we would stack rank it based on the information that you provided. And I was like, well, no, that doesn't make sense because here's a little bit more information. It's like, oh, well, thank you. Thanks for that additional information. Well, based on that, maybe it actually should be organized more like this. And it just gives you an example of like, is that really decision support? I don't know, but it's worthwhile. Like, it's providing you something that's helping you to refine your Thinking and make better decisions.
Juan Sequeda [00:22:45]:
Well, I think at the end of the day, you had, you started using it, there was a goal that you had in mind, right? I need to make a prioritization for these projects, right? I mean, you're trying to come create an outline for a talk that you're going to give. I'm doing the same. I do. I use it for my talks and stuff. And I say, I want you to challenge me. So I have these very bold claims I'm doing. I want you to go challenge me. And then I like, then I get follow up questions, how do you do this? And then I keep doing it. So I have a goal, right? My use case here is that I want to be challenged by somebody and I could get challenged by, hey, Tim, let's go to spend some time. But you know what, I'm like walking around thinking about this right now. I need to go talk to somebody right now. So that. So all this is, all this to say is that we're using this because we have a goal in mind, right? Which brings us back to like the use cases. Like, for us, we had a very clear use case. And I think sometimes it's interesting that we're having these discussions how we're using AI, but because we have our own use cases in mind right there. But then sometimes we go back to the organizations, we're like, oh, let's just go use AI. I'm like, no, but what was your use case that you wanted to go do? Then we kind of get like, we, we get lost. And I just find that fascinating, sadly, that we have to kind of say the obvious. What are the business questions you're trying to go answer where you like? It's kind of, I don't know, like, why does this get lost? And how do we get people to start thinking more again about the use cases?
Colleen Tartow [00:24:05]:
I mean, I think it's this tech first attitude that we tend to have, which is like, ooh, a cool new gadget. Let's figure out how to use it, right? And there's some gadgets, like an iPhone or whatever that are the most incredible gadgets ever. But remember when the Segway came out and they were like, it's gonna change everything? And it came out and now they're used for like, city tours. You know, I, I've never been on one. I would probably crack my skill if I went on one. But I, you know, it's like, it's a cool new gadget and we're figuring out how to use it, right? And I think that people are taking big bets on it. And there's a lot of companies up and coming that are helping people use it. Like, I see a ton of small startups right now that are like, in the space of helping operationalize AI and helping add the human into the loop, adding the judgment of humans into this automation so that you can then use it in a responsible way. And you're not completely automating things, but you can get back to focusing on the innovative part of it. You can get back to focusing on that creativity that we talked about. So that's sort of where I see us going right now. You know, that's why people are talking about agents and rag, because it's like most people aren't going to train a model, right? Like, nobody's going to train. Like there's like 20 people, 20 companies training models right now. And that's good. Like, you know, they're all slightly different and that's fine. But most people just want to do rag and just want to do, you know, some agent that makes sense for their business. And so there's a lot of companies coming out to help you do that, which I think is really cool.
Tim Gasper [00:25:40]:
Yeah, absolutely. Well, to kind of bring things back a little bit to streamlining data value. Right. I think we talked a little bit about the importance of use cases and kind of bringing it back to those use cases. I think we talked a little bit about, let's call it like, understanding the limitations of technology. Like, what is it actually good for, what is it not good for in terms of how you're going to apply it? What else would you point to as some key opportunities that we need to be focused on? You know, whether it's people process technology around streamlining data value.
Colleen Tartow [00:26:13]:
I think it's the people, right. Like, I think that it always comes back to the people. And, you know, I've been in leadership and in management for a really long time now, and it just, it always strikes me, you know, I have this like, quote revelation every couple of months where I'm like, oh, it's the people that are the problem, or it's the people that are making this work. And it's like, you think by now I would have learned, but I definitely, I see the people are the ones who are implementing the processes and the technologies. But, like, the people are the key between. They're the creativity, they're the innovation, and they're the conduit by which the data becomes the value. Right? Like, there's no way to automate an entire engineering data engineering pipeline. Right. Like people ask for that all the time. You can automate the technology part of it, but you can't have a machine automatically figure out how to curate your data because you're curious. Your creation algorithm wouldn't know what you meant by customer. Right. It would be like, well, do you mean somebody who's purchased your product or somebody who's looked at your website or somebody who's purchased more than one or you know, depending on what your business is. Like everything has context and so the people provide that context. So I think it always be, it always comes back to being a people problem, a communication problem or communication people success. Right. Like when you have the right processes and technologies, you can kind of let the people focus on the creativity and the success of the business.
Juan Sequeda [00:27:37]:
All right, we got so many T shirts right here. T shirts. So I want this one. Oh, it's the people that are the problem. And then probably in the back it says, oh, it's the people that make it work.
Colleen Tartow [00:27:48]:
I know, I try to put the positive.
Juan Sequeda [00:27:53]:
That it's both. Right. Like oh yeah, half glass full, half empty. But I love people provide the context. And I think that that's it. And look, we can't automate it, but we can definitely streamline it, make it more efficient to do that. And I think kind of tie this back to AI. Like that's how we should be figuring AI. Like look, I'm with you that the chat bots are the obvious thing that natural language, the SQL is. I remind people this is a problem people have been working on for 30 plus years is not a new thing. And we're just so, yeah, it's something it's not. We, we're kind of superpowering it. But what we realize is that the problem is the context and the people around that stuff. And, and it's just we, that's hard to scale. We got this great thing that can kind of help us, help us to scale a little bit. Help us scale. We should use that kind of. That's why I'm like a big push and advocate of focusing on what I call treating, treating knowledge as a first class citizen and, and doing all this context, this data knowledge work. And I'm not saying we got to automate. We're gonna, we can't automate and take the people out. I just want to superpower them saying we can do more of this faster and get more people involved.
Colleen Tartow [00:28:58]:
Yeah, yeah. Like, I mean a lot of the curation is just the application of like business Logic to the data. Right. Like, that's really what it is. And so that's what you want to allow people to do, because that's again, where the creativity and innovation comes in. So I think that's the focus for me, whether it's BI or AI or anything else. Right.
Juan Sequeda [00:29:18]:
I like how you, because you've used the word curation a lot and like how your definition here is connect the business logic to the data.
Colleen Tartow [00:29:27]:
Yeah.
Juan Sequeda [00:29:27]:
Like the very crisp and succinct definition right there.
Colleen Tartow [00:29:30]:
Yeah, yeah. It's like defining what. I always use the joke of, like, what a customer is. Right. But then Veronica Durkin gave a talk once, and at the very beginning, she was like, here's a piece of data for what does it mean? And it's like, I don't know, like. And she's like, okay, I told you, it's a day of the week. Is that enough information? And it's like, none of that, none of the data means anything until you apply the business knowledge to it. It turns out four is the day of the week. Well, is it Wednesday? Is it Thursday? Like, what's the.
Tim Gasper [00:29:58]:
What day does it start on? Does it start with zero?
Colleen Tartow [00:30:01]:
It could be Friday. It could be. It could be any day. Right. And so, you know, it's such a good example because it's like, you know, it could be any day and it means nothing until the person comes in and applies that specific businesses logic to it.
Juan Sequeda [00:30:16]:
Oh, man. This is. We're, we're. You know, I, I have the hope that with this is the moment that we're actually kind of realizing that we need to go back and, and kind of technology not set aside in the sense that, like, we know this is working and let's go invest into people and the context and this knowledge. I think people are realizing that little by little. I mean, last week I was on. I was on the road and that's. That was what I was talking about every day. And, and you can, you can feel it. I can just feel it now. Obviously, I'm very biased about this stuff. And then kind of, this is my confirmation bias. It'll acknowledge that. But I've also been talking about this stuff for almost 20 years. I'm like, not now.
Colleen Tartow [00:30:56]:
I'm.
Juan Sequeda [00:30:56]:
Now I'm hearing it more than. Definitely even five years ago and definitely more than 10, 15 years.
Colleen Tartow [00:31:00]:
Yeah. So, yeah, no, I mean, we saw all this whenever data science was a thing too. Too. Right. Like every college and grad program was pumping out data scientists and they all got hired and it's like, well, what are they going to do? Well, they're going to do data science. To do what with what data? Right. And it turned out that the data engineering was still the crux of the issue. So again, the people are the problem and the solution.
Juan Sequeda [00:31:25]:
No. So now that we're kind of talking a lot about people, I'm curious when it comes to the people and driving kind of streamlining that data value, what should be the. Are there any different changes or focuses in what we're seeing in roles today when it comes to data? Like is there new roles or consult. Because we talk about consolidation of tech stack and stuff. Is there something when it comes to roles and people in organizations that you're seeing?
Colleen Tartow [00:31:52]:
I am seeing more people consider data engineering, which makes me so happy. It feels so, so happy. And I seeing a lot more people involving data in their careers and ending up as data engineers coming from other disciplines, which is really fun for me. I mean I came from academia, right. Like I wasn't a software engineer and then data engineer. I didn't follow any like traditional path necessarily. I was a physicist and so, you know, but I was, I saw this woman online who was a librarian and she had a master's in library science and she started using data more and more and she ended up being a data engineer. And I just think that's so cool. Right. And so I think a lot of careers kind of lead down that path if you want to, which I think is really great. Now with AI, again, like I think maybe people realize this, maybe they don't but like 90% of the problem is still data engineering. Right. Is like pulling the data in making sure the data is curated enough for doing gen AI on it and running training or rag on it. Right. And so I think that, you know, people are excited about AI, but they're realizing that as well the data is still the key.
Juan Sequeda [00:33:10]:
One of the things that I've been pushing and then like I want to use as an opportunity to kind of get your feedback and either push me or kind of or let's move push together is I think it should be more because words matter. It's really data and knowledge work. Data, knowledge engineering. Because I think that the some I feel that data engineering ends up being kind of that term itself means more technical people thinking about oh, the data systems and the pipelines and that needs to be set up. Right. They are doing the curation of there but I think they're treating it, I mean that it's being. There's a Little bit of a separation when it comes to like connecting, understanding the business logic and stuff. And I think when I use a term, knowledge, for me it's more about connecting it to the make, making that, bridging that gap with the business and stuff. And, and I'm not saying. Well, and the, the unicorn can be people, somebody who can do the technical and the, and the, and the business side alt together. Because I mean talking to people and extracting that knowledge is, I mean that's, there's technical technicalities around that. But I think that we just, I want to emphasize more the word knowledge around this. Data knowledge, knowledge engineering. I'm curious what your, what your thoughts are about that.
Colleen Tartow [00:34:22]:
Yeah, absolutely. And I think, I mean to me this is why when I've run data engineering orgs, I've always pushed to have a product manager so that you have someone who's the voice of the customer. Right. Like the voice of the business, translating the business to the data engineers. And I think in some ways that's not really the right term. I mean data is a product, it is a data product manager. And I think maybe that that is the right term. I don't know. But I think, you know, if data is a product, which it should be, and AI is the product, you need somebody who's like representing the voice of the customer, whether it's internal or external. Yeah, I think, I think that without the context, what are, what are data engineers even building? Right. Like pipelines to nowhere. Right?
Tim Gasper [00:35:13]:
That's a good question.
Juan Sequeda [00:35:14]:
That's a good one. I'm just seeing here Korina's comment. Knowledge comes with experience of the data and data workers do not have that context. Yikes.
Colleen Tartow [00:35:24]:
Yeah, yikes indeed.
Juan Sequeda [00:35:26]:
Yeah, I really like you said this, who represents the voice of the customer, of the end user. Right. I think. And without that context, data goes nowhere.
Colleen Tartow [00:35:38]:
Yeah, it's like the pipeline, like what are you doing with it? You're just like putting it into a data warehouse. Like how are you curating it if you don't understand?
Juan Sequeda [00:35:45]:
Yeah.
Tim Gasper [00:35:45]:
And I just had, I had an image of my head of like a construction worker, like building a road. And they're like in the middle of the desert and you're like, what are you, what are you doing here? And they're like, ah, Ticket told me to do this.
Juan Sequeda [00:35:57]:
And the road has strong found technical foundations. Right. It's going well but. Yeah, where is it?
Colleen Tartow [00:36:02]:
Very well built road.
Juan Sequeda [00:36:04]:
Yeah. Well, so let's talk a little bit about like maturity levels. I'm curious about like how are you seeing, like, what are the different progressions of maturity and then when do you see things that are like, not well curated and examples of how you get better and better at this?
Colleen Tartow [00:36:23]:
Well, I think a lot of organizations have been focusing on data for a while now, which is the good news, right? Like, I don't think there are a lot of companies that are, you know, completely ignoring data unless they have a good reason. Right. And so that's probably, they're probably not doing great in that case. But I think most companies are to the point where they, at least on some level, are data driven. And so as they start to look into more complex use cases and complex technologies like Gen AI, they're going to build on what they already have, which is good. The challenge is they still need to know the use cases, they still need to know what they're going to do with AI. All of that still applies. But I think that we're in a good starting point in a lot of ways because at least folks are aware of what they have, right? They're aware of what data is out there, which is the first step, right? Admitting you have a data problem, it's the first step.
Juan Sequeda [00:37:19]:
Well then, so you admit you have the data problem, right? People realize it and then, and we're starting to go see the whole three days of product. What are you seeing right now that are like the North Stars that we should actually. The north exam, the North Star examples that people are refollowing in your experience.
Colleen Tartow [00:37:35]:
I mean, I think it has to drive revenue and if it doesn't, it was a nice experiment, right? And I think that's really what it comes down to because you can build the most beautiful data pipelines, you can build the coolest things in your app, and you can have awesome agents that are doing really cool things and translating SQL into natural language and vice versa. But like, if more people aren't using your application because of that, or more people aren't using or buying your product because of that, then it was a waste of money and time. And so I think that's what folks are struggling with right now is they're trying to find like, what are those things that are going to actually make more money, they're going to bring in more revenue with, with these new technologies, you know, in order to justify them because they are incredible, incredibly expensive, right?
Tim Gasper [00:38:24]:
Have you seen any use cases or are there any stories that you can tell that you've seen of whether it's your customers or like people in the space that you're like, oh, wow. They're doing something around data that either is interesting, that it's, in terms of it driving revenue, or, you know, or represents a solid level of maturity when it comes to all.
Colleen Tartow [00:38:45]:
Yeah, absolutely. And I'm going to go back to the pharma example. So the pharma example I love, and maybe it's because I'm, I live right near Cambridge, Mass. Where there's a ton of pharma companies doing a lot of this stuff. But they are able to use Genai to like, create new genomes and new. I mean, I'm not that kind of scientist, so I'm going to probably say this wrong, but new genomes, new kinds of medications that are customized to people based on their genetics and they can do it. They've. They were able to do that, I don't know, five years ago or something, but now they can do it that much faster. That makes it actually more of a reality. And so they can recognize more revenue, save more lives, you know, build all these medications faster and get them through the pipelines faster. And so I think that's just like my favorite example because it is really positive. It's like people are going to have a better quality of life because of this. Now that said, you know, there's the negative side of it too, that a lot of people are going to not have access to that, etc. It's using a lot of energy. Like, there's all these downsides to it. But I think, you know, getting back to the question of maturity, I think those companies tended to be in a more mature space to start with. They have a lot of structured data, they have a really good handle on their data, and they tended to be able to move faster once they got access to these, you know, J algorithms and LLMs, ETC.
Juan Sequeda [00:40:09]:
You know, one of the, one of the things I'm kind of going through our notes and I'm seeing like the main takeaway of this conversation that I'm having is creativity. And I think creativity is something that drives us to be innovative, to go do things that we didn't do different, that were. Do things differently that can help us drive kind of better outcomes in our, in our business and stuff. We should be using AI to help us kind of foster that creativity and just know there's tools around that stuff. And I think what, what, what the farm example you said is like, oh, you're using this to be, find greater ways to do this faster. I mean, they were doing it before, but we're doing it faster today. I mean, our conversation of how we were using AI ourselves kind of internally, it's like we just need to be more creative kind of to streamline the way I need to come up with, with an outline faster. Right. How do I give, I need to, how to prioritize the stuff I want to go do. Like, I think that's kind of like the main takeaway for me about this conversation is that we have the opportunity to be creative. Creative. We should be using these technologies to be creative and that's what's really going to make us be innovative. And I love this. It has to drive revenue otherwise. It was a nice experiment.
Juan Sequeda [00:41:19]:
But that's the honest, no BS statement right there. Just, just before we head to our lightning round, just kind of to close it, close this a little bit on, on, on, on the stack. Where does. We've talked about the data set, we've talked about AI. What are your thoughts or how you seeing the technology stacks between data and AI? Are they separate, coming together? How does this look like?
Colleen Tartow [00:41:39]:
Oh, they're definitely coming together. I mean, I think as people understand that they don't need to train a model in order to use this, they don't need a GPU to do a lot of this stuff. And so I think it's becoming more accessible, which is great. There's these AI cloud service providers out there that are making it more accessible. If you do need GPUs, you can like rent them and things like that. And I think that it's just going to get better in terms of energy efficiency. That's the thing that really bugs me, honestly, is the lack of energy efficiency here and the incredible resources that we use to do. You know, I've heard things like every time you run something in ChatGPT, you're dumping out a bottle of water essentially and it's like, that's horrifying. Right. But I think we're getting a lot better and hopefully quickly. Yeah, I, I mean I think that we're, we're definitely getting better and I'm excited to see where this goes. I mean, I think, you know, we're a creative race, right. Humans usually.
Juan Sequeda [00:42:40]:
I am of the, My position is I have faith in humanity. You see his look at history.
Colleen Tartow [00:42:51]:
Yeah.
Juan Sequeda [00:42:52]:
Humans are very resilient and so. Wow. So much stuff. Let's hit our lightning round questions, which I see that Tim. So just people who know, like we have our, our doc and Tim usually writes all the lightning round questions and I'm seeing them for the first Time right now. So here we go, people. Process strategy, technology. Which one is the biggest bottleneck to streamlining data about people?
Colleen Tartow [00:43:19]:
Always. We're going to get the T shirt, remember?
Tim Gasper [00:43:22]:
Yeah, we got people are the problem, people are the solution. Depend which, which one's going to be on which side is the question. Which one goes in front? All right, number two question. Is the AI stack part of the data stack?
Colleen Tartow [00:43:37]:
It should be. It should be. But your BI stack can't do AI typically. So you need to start rethinking things and that's the big challenge because it's expensive to do that and you have to prove the value, etc. Etc. It's like a chicken and an egg thing.
Tim Gasper [00:43:52]:
So slight bonus question here. Do you see that people are approaching this a little bit like separately, but, but you see them converging. Is that kind of what you see happening?
Colleen Tartow [00:44:00]:
Yeah, yeah. A lot of enterprises are doing that. They're keeping their legacy, sort of bi world that they already have up and running over here, not touching it. Figuring out the AI and then the longer term plan is to consolidate.
Juan Sequeda [00:44:13]:
All right, bonus, bonus question on this one. Curious, from your perspective, were you seeing kind of decisions being made around this from a reporting standpoint? CIO, CTO, CDOs, like who's getting, who's buying and owning these, these stacks and these projects. It's typically AI and data between AI.
Colleen Tartow [00:44:33]:
Yeah, like, I mean typically there is like some sort of person in charge, like a VP of AI or something like that that you would, that would own the initiative and the strategy on how to implement it. But it is typically under the CTO is what I'm seeing, or the cdo, depending on how that structure is. I mean some places you've got a CDO reporting to a cfo and so it's like this tends to be more on the technology side to start.
Juan Sequeda [00:44:56]:
Okay, next question. So right now very few applications have been deployed to production. Do you think that will change as the market matures or is there still a bigger piece that is missing?
Colleen Tartow [00:45:08]:
It will absolutely change and accelerate.
Tim Gasper [00:45:12]:
Awesome. And last lightning round question. Do you see data and AI leaders caring more about governance as we go forward?
Colleen Tartow [00:45:24]:
I mean caring, yes. Knowing what to do to implement it? Not really. We're still figuring that out. Right. As a society. Right. Like governance is always the last thing. So it's like we'll figure it out, we'll get it running and then we'll figure out the governance.
Juan Sequeda [00:45:40]:
But, but it's interesting that you say I, I, I agree that that's kind of what happens. But then we are still talking now about like this thing. This is lack of transparency. We need to have responsibility and stuff. Like, doesn't governance all there? Or are we just gonna like, say it and then like, screw it? It like, this is not responsible.
Colleen Tartow [00:45:57]:
The monster. Then you figure out how to kill the monster.
Juan Sequeda [00:46:01]:
But we're still kind of doing that with data today, right?
Colleen Tartow [00:46:04]:
Oh, totally.
Tim Gasper [00:46:07]:
That's funny. That's funny. I feel like a lot of conversations that we have even at Data World around AI governance, for example, can be a bit circular, just to be honest, where it's like, hey, so what are you guys trying to do around AI governance? Well, you know what, what does your product do around AI governance? Well, it's like, well, first tell me about kind of what your strategy is around AI governance. Well, what should we be doing?
Juan Sequeda [00:46:40]:
I'm gonna take a risk. I'm gonna take a risk here. Say this live. But I'm like, when people come and say, so why do you need a data catalog? Oh, well, because we can't find data. I'm like, yeah, no shit. Is there a. Like, I know everybody has that problem. What are you trying to go do? Oh, we're trying to democratize it. Like, I know. What are you trying. We're trying to be data. I know. What are you. What did that. Sorry. So if you feel, if that resonated with you, that you're that person. So come on, we need to get. We get to move. Streamline data value. Listen to this episode here. All right, takeaway time. Tim, take us away.
Tim Gasper [00:47:17]:
All right, well, Colleen, awesome, awesome session today. We started off by asking the question. Question, honest, no bs. What does it mean to streamline data value? And you said that it really is around a couple of things. One is around consolidating the data stack. Really trying to simplify, try to combat sort of, you know, I don't think you use this phrase directly, but I could hear you kind of alluding to the modern data stack kind of craze of like millions of things. You know, Matt Turk's big, big, big diagram of all the different vendors and things like that. You know, simplify, shorten the path and focus more on the end value. Really focus on the use cases that you're trying to drive. You know, technology is really a means to this end, which is value for the organization. And you mentioned that when it comes to AI, everyone is kind of, you Know, not being creative enough as they think about the right ways to apply it in their organizations. You know, people immediately think to things like a chatbot, or they think to things like, oh, I want to use natural learning or natural language to actually generate SQL. What is the value of those things? There are certain use cases certainly where those things can be super useful. Perhaps the chatbots more than the natural language SQL generation, because they in themselves is not enough. You have to tie it into a bigger problem that you are trying to solve. And, and you mentioned that, like, you know, there are way more creative ways to leverage these technologies that are maybe a little bit more specific to your company and what and how it is that you make money. Like, for example, if you're a pharmaceutical company, the question you should be asking is, how can we use data and AI to help us, you know, make better drugs, more effective drugs, bring those drugs to market faster. You mentioned later, you know, customized drugs as an, as a very creative example using AI. It's hard to force creativity, though. And then you mentioned that's why diversity is really important. And, you know, you gotta, you know, think about what you can, what you can do now versus what you can do later. Right. And you know, there's going to be simpler use cases, there's going to be more complex ones. Yeah. What are your thoughts of AI and the role that it plays in decision support? So thank you and shout out to Bill for a great question there. You said it can't be fully autonomous. You need to have critical thinking, thinking. What is different today is the lack of transparency, which can be a huge challenge. It's good that if you're nervous about it, you know, but you can't write it off. You have to, you have to address that. And it's a cool new gadget and they're trying to figure out how to use it. And, you know, you got to figure out how you can keep a human in the loop so that we're not, we're not just giving the farm of all the decision making away to the robot, because that's probably not a safe path forward.
Colleen Tartow [00:50:05]:
Absolutely.
Tim Gasper [00:50:05]:
But so much more. Juan, what about you? What were your takeaways?
Juan Sequeda [00:50:08]:
I love. I like that we talked about the people. Right. It all comes back to people. The T shirts. Oh, people are part of the problem. Oh, it's. People are making this work. So I guess people are the solution or people are the key to the creativity, to the innovation. I mean, we can automate that tech part, but we can't automate that Curation part, because that's the meaning. Because at the end of the day people are, are the ones who provide that context. And by curation we mean I like this connect the business logic to the data. And one of the things that we've seen is like a lot of big focus on kind of shift more towards data engineering. And we get this conversation about how I see that we need to have more the data knowledge engineering. And I think I really love how you brought this up. It's like who's representing the voice of the customer, of the user? I think that's a really good question that everybody needs to be asking. Who's working with data? Like who's representing this? Just like a kind of a quick side note us, the data world, we always have like that, the, that we call it the, the. Well when we had our office, every, every room had a blue chair and it was always representing the voice of the customers. Like the customer is sitting in that chair. So whenever we're having a meeting we want to say okay, this meeting, how is it helping the customer? And we always had that blue chair in there. And then we actually, I mean team team's our chief customer officer, we talk about, we have a, the blue chair award where we're like, we give it to, to, to people, employees, data world who are not in the custom or organization who are doing, doing so much for our customer. There, there's the blue chair.
Colleen Tartow [00:51:30]:
I love it.
Juan Sequeda [00:51:31]:
Yeah, so it's a, that's why I really connected with this. Who represents the voice of the customer? This is super, super important. And I think without context, data engineering, you're just building pipelines to nowhere. Another T shirt code. If it, it has to drive, it has to drive revenue. Otherwise it was a nice experiment. If people are not buying more of your product, whatever, you're kind of a waste of time right there. So you gave another pharma example, right? You're using gen AI. They're building new drugs that can customize people's genetics and they're just doing it faster today. Like that's an example of how we can use these technologies to drive revenue. And quickly we wrapped up with the tech stack, right? The data and AI tech stick are coming together. The cloud platforms and vendors in general are making this very much more accessible so you don't have to buy all these GPUs. And very importantly, we're getting better at the energy of efficiency. So. And then finally kind of just the takeaway that wants takeaway of the takeaway. Everything is A way to streamline data value is through creativity, and that's people. And we can use that technology to make us more creative so we can streamline data value. All right, how did we do? Anything we missed?
Colleen Tartow [00:52:35]:
No, I think you got it all. That was great.
Juan Sequeda [00:52:37]:
Awesome. Well, wrap it up. Three questions to you. What's your advice? Who should we invite next and what resources do you follow?
Colleen Tartow [00:52:47]:
Oof. Okay, what's my advice about data engineering?
Juan Sequeda [00:52:53]:
How about data about life? By the way, I say this all the time because I'll never forget somebody. One of our guests, long time ago, Pat Berry. So his advice. Have a very long USB cable.
Colleen Tartow [00:53:06]:
Oh, that's good.
Juan Sequeda [00:53:07]:
I bought that. And I always have whenever I travel. So. So about life, data and data and. Or life.
Colleen Tartow [00:53:13]:
All right. My. My advice is something that I told my child who just got his first laptop. Anything you say online can be screenshotted and shared. Do not forget that ever. Nothing you say online, whether you think it's a private text to someone, everything can be screenshotted and shared.
Juan Sequeda [00:53:34]:
That's fantastic.
Tim Gasper [00:53:35]:
Very true.
Colleen Tartow [00:53:39]:
Okay, so who should you have on. Have. I don't think you've had Lindsay Murphy on. I would say Lindsay Murphy also. I like nerding out with Monica Miller is amazing. I think she's actually. I think I saw her make a comment.
Tim Gasper [00:53:54]:
Yeah, she actually left the comment. That's awesome.
Colleen Tartow [00:53:56]:
Yeah, she's awesome. Who else? Selena Wong is really amazing. She is the CEO of Data Culture. She sees a lot of different industries through her consulting work. Yeah, there's three for you. Perfect.
Juan Sequeda [00:54:12]:
Love this.
Tim Gasper [00:54:12]:
Love it. Thank you.
Juan Sequeda [00:54:14]:
And so I've been following Lindsay, so definitely. So, Lindsay, if you're listening here, we can say this episode. Selena, I've. I've met her in passing that at a. At the conferences Agriculture and Monica, if you're listening, we'll reach out to you soon. So cool.
Colleen Tartow [00:54:29]:
Awesome.
Juan Sequeda [00:54:30]:
All right, and then what resources do you follow?
Colleen Tartow [00:54:35]:
I mean, I love Joe Reese's stuff, Ben Stansel stuff. Like, they're always just hilarious too. Right. They're always entertaining, which I appreciate. I like following some industry analysts too. Like, Sanjeev Mohan is really good because he always has, like, a good pulse on, like, what's the new hotness? But it's very realistic. Take, like, you know, is it really hot? See who else.
Juan Sequeda [00:55:01]:
Sanjeev, I call it. He's the. He's a true, honest, no BS analyst.
Tim Gasper [00:55:05]:
He is.
Juan Sequeda [00:55:06]:
He love. I mean, I'm so. Yeah, I'm gonna text him right now that we're talking about him? Yes, I love him. And what I love about Sanjeev, you know, people, if you don't know from Sanjeev, like, you are totally missing out. Like he, he, he, he's a guy, when you're, when, whenever with him at a conferences, he is like freaking taking notes every, all the time. And when he writes these posts, they're incredibly thoughtful, very dedicated and, and he's not. There's some analysts who are like high level, fluffy. Like this guy can go all the way top to all the way to the details and challenge you.
Colleen Tartow [00:55:40]:
So yes, absolutely. Love Sanjeev. Cool.
Juan Sequeda [00:55:46]:
Well, very quickly, next week we have Andrea Gira, who, if you've been following him on LinkedIn, is all talking kind of a lot of my stuff too, which I'm interesting. I'm excited all about knowledge and focusing on knowledge and semantic stuff. That's our conversation next week. And with that, Colleen, thank you. Thank you, thank you so much as always. Thanks to Data World, who lets us do this every single Wednesday. And we've been doing it for on our fifth year. And Colleen, thank you so much.
Colleen Tartow [00:56:09]:
Thanks.
Tim Gasper [00:56:10]:
Cheers.
Special guests
