Data Chaos Bits: 🔮 Unveiling the Future: Insights from Snowflake Summit 2023

0:00:01 - Tyler Wells
All right, welcome to the data chaos podcast. This is going to be a special edition data chaos bits edition, where we recap the Snowflake Summit from 2023. Two weeks ago today, actually, all three of us that's myself, Nico Acosta and Dave Ganley We're all at the Snowflake Summit out in Las Vegas And, aside from the unbelievably high production value that that crew put on, we had a ton of takeaways from that conference. We spent three days learning all about what Snowflake has to offer in their ecosystem, all of the new things they're building, all the cool stuff that their customers are doing, everything around AI, LLMs and everything else, And we wanted to have a chance to sit down and talk about it. Guys, what did you think?

0:00:53 - Nico Acosta
Yeah, I think one of the things that's shocked out to me was that Snowflake clearly had a hidden agenda. They wanted to get us fit for the summer. I think we walked 12 miles a day between the Caesar Forum and the palace.

0:01:13 - Tyler Wells
My watch had for a three day total three day total of 31.2 miles walked.

0:01:19 - Nico Acosta
That's pretty incredible. But, fitness aside, one of the things that I think is a game changer are dynamic tables. They announced them last year as a private preview, they announced them this year as a public preview And they really simplify the transformation workloads. If I had to guess, they're gonna be really eating into the transformation tools like DBT. It is just so simple to set up a dynamic table based on a SQL query and just transform your data continuously. A lot of the feature that they have where you define the lag And you say, hey, i want a one minute lag or a 60 minute lag, it's geared towards that SLA that you want to provide with your data And so very, very excited about that functionality, very excited about being able to chain these dynamic tables together And, yeah, really, really looking forward to seeing where that goes.

0:02:28 - David Ganly
You know it was an important release simply because the LinkedIn, the Reddit chatter, the debates were immediately raging about like would you use dynamic tables versus DBT? And there's a lot of good use cases for continuing to use DBT for different tooling reasons and all sorts, But to your point, it's hard to see how a bunch of workloads want to immediately be easier to grok, understand and implement if you just use dynamic tables. If you're already doing everything else inside Snowflake, then dynamic tables could be an obvious choice for your transformation.

0:03:01 - Tyler Wells
Well, there's a really I was gonna say, think of, we switched over, right, and all we did was dynamic tables. I mean that cuts out kind of an entire service that we have to run today, that we're running in Fargate And you know we get everything that we need, all the benefit of, you know, being able to decorate those tables, expand those tables out, and really wouldn't change our workflow much other than we kind of cut out an entire service.

0:03:25 - Nico Acosta
Yeah, that's right, and I attended the talk by two of the engineers that built dynamic tables And they went into the detail on how they handle the incremental updates. It's pretty amazing And it's a worthwhile watch when those talks come out.

0:03:51 - Tyler Wells
Now the talks. I mean the whole talk, kind of like they had so many of them set up. I mean it went everything from hey, here's the intro, to kind of Snowflake, which was pretty basic, to hey, let's build APIs on top of your Snowflake, to the inner workings of dynamic tables. I mean again, the value you get out of that conference I felt was very high. I definitely walked away After not going to a conference. I think the last one I went to was definitely a Twilio signal many years ago And then obviously COVID and getting to go back to that. That was a really nice entry point back into the conference world. So definitely enjoyed it. I think the other thing that was talked about heavily was LOMs and AI and the partnerships that they now announced with NVIDIA. I know on Monday they did a fireside chat with the two CEOs of Snowflake and NVIDIA. That was obviously well attended and definitely got a lot of fanfare. Dave, what did you see there? Take away from sort of the LLM and AI side of the house.

0:04:53 - David Ganly
Yeah, it's impossible to do a conference right now without speaking heavily about AI and everything else. But obviously, for Snowflake, the CEO made it really clear that the whole strategy is if the data is inside Snowflake, then it makes it obvious that that's where you're gonna do your machine learning, that's where you're gonna build your large models, and the NVIDIA partnership is a big kind of foundational piece. I mean, that's what the framework does right, it's the foundational models, it's the GPUs that you need to build things with reasonable speed. So that's a key pillar. We also saw kind of the more kind of DevTool, kind of WizzBang demos as well.

They did, you know they did the when I loved was the kind of GitHub co-pilot type style where you're in the SQL editor, you drop a comment saying you know, give me sales over the last year for each city, and then it spits out the generated SQL and then all you have to do, hopefully, is just hit, run or maybe tweak it. The other one that was super interesting was the document AI, the kind of unstructured thing I think they acquired. They spoke about this in the keynote. They acquired a startup, i think it was called. I wrote a demo notes, aplica and kind of analyzing structured data. So the demo they showed was super impressive, right? So a whole bunch of PDFs are literally scanned and uploaded and stored, and then they were running ML models to literally either run queries and then inspect the PDFs fast enough and then return that or just process all of the data in the PDFs ahead of time and bring it into Snowflake tables.

So there was a lot on display and I think, yeah, the kind of strategic premise is completely sound. Right, you already are making Snowflake your source of truth. You're ingesting tons of disparate types of data into Snowflake, so where are you best placed to build your models, instead of going out to a whole bunch of third party systems, whole bunch of third party tools, the more of that they can enable on top of Snowflake, so you can actually build, train, get everything set up. It makes perfect sense. So inevitable, but also like super exciting and an awful lot of use cases unlocked with what they announced.

0:07:19 - Tyler Wells
Yeah, i've not thought those talking tracks that they presented were enviable and the fact that there was just some really cool tech, you know, being displayed, that you know I'm sure it's not here quite yet. Most of the things we always see at conferences are kind of you know, a little bit in the future not too far, but if it's as close as it was that we saw, looking forward to seeing those come out.

0:07:41 - David Ganly
Yeah, what I loved about it just to kind of really put a finer point on it it's just just the way they presented it is like you are on a consolidation journey. You are again in your data into snowflakes snowflakes a great place to land your data. You're going to get to this step. We know you are. We know you're all going to be building these models.

We know you're super interested And so it just fits with what people who attended the conference were already thinking. It's what their bosses were asking them about, right, so they can go back now and say, yep, you're getting on a snowflake. That being our source of truth is a is a solid call, because we know all this. You know ML training, work, workloads and use cases and tooling is coming to snowflake itself, so it's a good bet for their own futures.

0:08:24 - Tyler Wells
And then the other narrative that was weaved pretty much throughout the two days of the big keynotes was data apps and the huge prevalence of data apps being built and visualized using streamlet. I feel like that was in a whole bunch of the demos. It was definitely in that big demo where they they brought on the snowflake employees to be actors, which was amazing. I thought they did a phenomenal job, but that was really cool because they were actually solving problems that we probably all solved at one point or had to deal with in our own careers. And then acting it out live on stage and then building a full solution in streamlet was pretty amazing.

0:09:05 - Nico Acosta
Yeah, and if you, if you play that out, you fast forward a couple of years, right, and you say where, where can this go? You could imagine a world where BI tools are built by data teams, right, and that they're they're smaller apps, they're purpose specific. Right, there may be like a small app. They're built by data teams. Right. There may be like a small like budgeting app or a small like sales compensation calculator or a headcount whatever, right, or a sales report that that are just built by the data team. The data team already knows Python. They might not have front end skills, but streamlet solves that. Right, it's built on top of, on top of snowflake. And one of the things that you, you, they mentioned that they think they didn't quite make it was the hosting of a streamlet apps. They were probably pushing really hard to get that, but you know it's, it's coming right. Then, once that's there, you just give a user access to an app, right?

0:10:25 - Tyler Wells
I thought about that. I think they actually they talked about it even being easier than that once it's hosted. So I thought one of the things that they had done was they built the app, they had the visualization and streamlet and then they wanted to share it by sending a link. But I thought it was, i thought it was that simple, like that. I thought that was insanely powerful because you could think of the the ad hoc question coming in about the data from somewhere high above the, the data engineer quickly, you know, writing the sequel to do all that stuff in their notebook, visualizing it in streamlet and going share to you know CEO at company xcom. Boom, there it is And I thought that was pretty powerful.

0:11:00 - Nico Acosta
That was very powerful And you contrast that with kind of buying like an old school traditional BI tool that is paid by the seat, that it has like these kind of clunky dashboards. I don't know, if I had, to guess I would bet more on like very granular, purpose built apps that are shared, that solve a specific thing And, yeah, the interesting being a ton of them.

0:11:39 - David Ganly
The penny drop in the demos, that the live acted, as you say, tyler, the stage show or the, the the theater presentation was the bit where, yeah, they brought the streamlet app back to the stakeholder but along with UI controls so that they could adjust it. So it's not just sending the link to the CEO and CEO gets the graph of, you know last week sales, but they also get some level of control. And that's why, to Nico's point, that the spoke app, where you're giving the end user who gets that link, the, not just a static graph that gets rendered or they have to go edit some sequel to change that, but you're giving them some level of extra control and user interface so that they can quickly edit it without any sequel, touching any sequel at all. So, yeah, seems like it unlocks kind of a all the way through to the front end in a very modern way but for a data data team who don't necessarily have those skills at all to kind of get all those components working together. So that's really nice.

0:12:41 - Tyler Wells
I know, watching all of that, like I got excited for, you know, the, the BI teams, the data engineering teams But I was a little bit I don't know if I want to say the word sadden, but I was. I was more like okay, well, what about the customer? And you know, you've got all these very cool tools inside and all these new capabilities, unlocking it for a lot of internal use cases, but I still felt the one thing that was lacking is how do I take my snowflake data and unlock that data, build apps on top of that and ship that to my actual customer in the form of like customer facing analytics and dashboards and anything else like that. That's just. I still don't think they're there yet.

0:13:21 - Nico Acosta
Yeah. So when you look at it, i think we all attended all the sessions about the SQL API and the Snowflakes API. I think a couple of takeaways. They were all packed right And they were packed with people like with enterprise folks, with software company people and looking to empower other teams in their organization to build with data right And to expose that data in their customer-facing web and mobile labs right. And what was clear from those talks is that there's still a lot of things that you need to do. They even recommended against using the Snowflake grants and authentication methods for multi-tenant customer-facing use cases because you still need to build a ton of logic and you don't have fine-grained control over the access. They recommended selling a cache to support high concurrency and reduce the latency. They recommended adding an intermediate service And you could see it in the audience just like gosh. The API is very simple but that's a lot of things I need to do to be able to have an API for the rest of the teams in the org to use the data in customer-facing sags.

0:14:49 - Tyler Wells
No, i feel like I felt that pain as we all attended Brad Culberson's talk on building a REST API over your Snowflake data. He's a part of the field CTO office. He's a principal architect there And he gave a two-hour presentation on how to build that And I swear we sat through the first 30-plus minutes of that of how hard it was to do. He went through all of the steps of all the things you need to do. He talked about Cognito, he talked about AWS API Gateway. He talked about Flask. He talked about getting your routes correct, getting permissions correct. I mean, there was just so much stuff And you had to do that. Most of that was all outside of Snowflake that you had to do. It was a pretty heavy lift to get all of that done. I was an extremely well-attended talk And you could see there's huge appetite for this And it was great to see, but at the same time, it's not an easy thing to accomplish.

0:15:42 - David Ganly
Yeah, the way I thought about that was it was a great talk, it was a great presentation And it was a great hands-on lab. They had a quick start, ready for folks to get started. People were working directly on their laptops. It kind of reminded me of the Twilio Superclass demos we used to do back in the day.

0:15:58 - Nico Acosta
It was really good.

0:15:59 - David Ganly
And I thought about the folks, the developers, who came out of that class with that prototype sitting on their laptop And they bring that to their CTO, their director of engineering, their product people, and they say check it out, here's a real API. I did the Flask setup. I got some of the authentication stuff working. There's basic methods And they go this is great, this is customer-facing analytics, this is something we need. And then they go, yeah, ok, well, what does it take to put that in production? And they're ah, there's a lot to do. There is a lot to do, and that's what the first 30 minutes or so, like you said, tata was about.

It was like there's a lot here that you need to consider. There's a lot of technologies to choose. There's a lot of trade-offs to consider about how you're going to do authentication, how you're going to model your data, how you're going to present it to customers, all the way up to the UIs. So, yeah, i just think of that conversation When you bring that demo. It's a great demo, it's great starting points, but you've actually still got quite a sizable gap to bring that, whereas some of the other snowflake enhancements were very much like oh, that just makes my workload easier. My day is now easier. The second that gets to a public preview or to a GA moment. That feature makes me X% faster. That API stuff was more a starting point, but there's still a lot of work to do. So it was clear that people were going to walk out of that thing with a great demo, but they needed quite a few more steps to get where they needed to go.

0:17:30 - Tyler Wells
Thought of doubt. So let's switch gears. Favorite happy hour that you attended Which one was it? Because there were a shit ton of those. Oh my God, there were too many to go to. That was crazy.

0:17:44 - David Ganly
Nika.

0:17:47 - Nico Acosta
Which was the favorite. I actually really liked the DBT, ph data one and the rooftop Beer park. Beer park, that Paris area, that was a nice setting, very well attended. It's refreshing to be outdoor.

0:18:08 - Tyler Wells
Yeah, i thought that one was packed. It was good, it was super well attended, but I think outdoors in that venue it wasn't super hot. Sure, it was in the 90s, but it was still a great day. I enjoyed that one as well, but the one that we were at previous to that I felt we had a lot of really good conversations until it just got too packed. Yeah.

0:18:29 - David Ganly
Because then you couldn't hear anything.

0:18:30 - Tyler Wells
It was insane in there The decibel levels. everybody's screaming at each other. You could only talk about things so much until you're just kind of staring at each other drinking beer.

0:18:42 - David Ganly
Yeah, the beer park one. That happened to me when I did. If you noticed, i snuck off to the. I'm thinking about the right one, the giant chess set in the corner.

0:18:50 - Tyler Wells
Yes yes, that's right.

0:18:52 - David Ganly
I'm sure that there's a lot of overlap between very talented data engineering type folks and people who are confident about their chess playing, so that was pretty fun. I don't know where to go with that, but I don't know if that's another point, it's a good takeaway.

0:19:06 - Tyler Wells
We need to Well, i think you escaped that conversation well, so things could have happened differently, but glad you got it out of there.

0:19:14 - David Ganly
Honestly, all credit to the folks who sponsored those. They were well put together. Just the basics of getting people into the door checking people's badges It was already well done. Really nice people from I think we went to. We met a lot of folks from census. They're just genuinely lovely people chat too. Yeah, all kudos to folks putting on those events. It was really well done.

0:19:37 - Tyler Wells
They were very well done. Definitely appreciated those as well. To get out of the expo hall and go check those things out.

0:19:46 - David Ganly
They were a great door one. They managed somehow to keep us all reasonably cool, both with drinks, which was lovely, but also just like blasting us with cold water. The Vegas heat was a lot.

0:20:00 - Tyler Wells
So any regrets, nico, on having Not having a booth, versus the gorilla marketing approach that we took of walking around with our I hopefully are now famous shirts that said you know ask me about my API and GraphQL, your sequel or Snowflake.

0:20:20 - Nico Acosta
No regrets. I think it is a Especially the smaller booths. I would say it's a toss up. If you get a good location on the small booth, you're golden. If you get a bad location, it's kind of worthless. If you're towards the back, nobody's walking there. It's really hard to be in the flow of people. So some It's my takeaway that it was that small booths were a toss up, whether they were valuable or not, highly dependent on location. The media booths seemed to be all very well located. You could see I think Twilio killed it with their booth. They had a line all the time, really strong play there. That was a medium booth they did. They were able to execute impeccably.

0:21:23 - Tyler Wells
Yeah, you seemed to like the medium booths to get value, high value and high draw. You had to have some sort of like a gimmick. right, it had to be a gimmick other than just coming to talk about the tech. I mean, obviously they're all there to talk about tech. I like what the Twilio segment folks did. You had to sign up and answer a few questions and then you've got like a laser etched water bottle, which is really cool. They had that line out there the whole time. They told us they got a little bit of trouble from the Snowflake people. We're kind of policing them because they're blocking the aisles and everything like that. But if you're Twilio segment, you're pretty happy about that. You've done well.

0:22:02 - David Ganly
Yeah, i would run through all of the cool booth gimmicks. I saw The nice kind of things people are hanging out. I'm nervous, though, because if I mentioned the things I thought were the best, i may not also remember the company, because some of the things were so distracting and cool that you kind of went what's this and what's going on? But to the meat of your question, tyler it was great to meet so many partners. It was great. So I felt like being mobile, walking through, doing a balance between like hey, i'll go check out that quadrant of booths and then hey, i've got a talk in an hour, i want to get to.

It worked really well, except for, maybe, the aforementioned hike between Forum and Palace. But I think for us getting the T-shirts, i've got to highlight that. I don't know how many you know. I mentioned a potential customer just stopping me and saying, oh yeah, okay, well, tell me then, right. So you know, tell me about the API, tell me what's. You know, just literally pointing at the T-shirt and saying, cool, explain it, which is exactly what we wanted. So I think that was fun.

0:23:14 - Tyler Wells
Yeah, i definitely had bites on both sides of the T-shirt. So we were Nico and I were in the big lunch tent and there was a gentleman walking in front of me. All of a sudden he stops, turned around and goes all right, i'll ask about it, tell me about your API. And then that was like a 15 minute discussion after that and handed out a few stickers, and that was great. And then the other one I was standing in line waiting to go into Brad's API talk and a woman walked by me and said I love GraphQL. She's like GraphQL over my snowflake. That'd be amazing. And so she got some stickers as well. She was very excited about it. So, yeah, it definitely was nice to stand out and have some good, some marketing slogans on our shirts.

Back to the tech. I think my the thing that I walked away with that I'm most excited about from a potential standpoint is the Snowpark Container Services. I think those have the potential to be a game changer for a lot of people bringing the applications to the data versus bringing the data to the apps and sort of flipping that whole thing on its head. I look forward to hopefully getting early access to that so we can start to build and see how our services and everything that we've built that is currently running in AWS and is already containerized. How fast can we sort of spin up our API and the platform that we need and get that right next to a customer's data?

0:24:49 - Nico Acosta
Yeah, that's, that was one of the big themes. Right, bringing the apps to the data, and it's a really powerful idea. Right, it applies to all sorts of apps. It applies to AI and ML And I think, when you kind of distill, that is, the data has a really strong gravitational pull right, where, if you think about it, there's a family of apps that need to be moved where the data is right. And, for example, the R platform, our Propell API platform. I guess you said we're eager to play around with the Snowflake containers and deploy them so that enterprises can deploy the Propell API layer on top of their Snowflake data within kind of the boundaries of their Snowflake account. Right, yeah. And then for everything else, right, there are. There are apps that you cannot move to the data, right. Imagine, like Shopify, shopify hits, it's something, or Salesforce, or all these. For everything else, there's an API, right, so it's, it's a, it's a way to unlock new, new workloads, right.

0:26:23 - Tyler Wells
Do you think so? if you imagine, imagine if you had taken that route from the very beginning, because that existed. Would you even have to do sock to? You could almost you could almost be like.

0:26:34 - David Ganly
I don't need it.

0:26:35 - Tyler Wells
What do I need it for? Right, i don't touch your data. It's over here, it's in your own, it's in your own. you know your own instance, your own, your own cloud. more or less?

0:26:44 - Nico Acosta
Yeah, probably not you and then a lot of the in, a lot of the compliance, and then just the fact that, like you can apply all the same, like snowflake policies, say, like the Access policies for IPs, right, it's like All of that applies to your snow park containers, so like you're within the walls of the organization and that's that's really a game changer with a very say, with a very Simplified operational model. It's not like, hey, this like super complicated Kubernetes cluster that you'll need to engineer, to maintain and to monitor. Like they've done a nice job Simplifying that. And when you look at the container services, like they, they build down the services, the jobs and the service functions Right. So like you can deploy kind of like far bit type jobs, you can deploy lambda type workloads, you can deploy continuously running services, and So very excited about those, those possibilities.

0:27:56 - Tyler Wells
Yeah, i'm super interested if they're gonna take sort of you know, a declarative language approach to You know orchestrating those services and what that looks like. I heard some mention of like a sequel, like, yeah, i.

0:28:11 - Nico Acosta
They demo with it. A sequel like kind of like, create a great service with all your parameters with, but I think that the core building blocks are an image registry that you pull from right and and then the, the manifest that describes what to do. So, yes, it is sequel, but there is some a a Kubernetes like YAML that you can't jam it all. Sequel that you need that manifest file.

0:28:49 - David Ganly
And just a nerd egg for a while. Right, like we were at Tulio, we all did developer experience at first. Various times we all worked on those kind of problems. It's really nice to see you know the evolution of the platform approaches. So you know, you think back to Salesforce being then no software paradigm and all the different ways that they approach the like. Let's get people's apps onto Salesforce or connected to Salesforce and now we're seeing, you know, a modern public companies approach to okay, well, how can we bring those apps on to our platform? And I just love that.

It's, yeah, it's containers. You can run what you want. We can run it in a, you know, a walled garden. We can have security and compliance taken care of. We can simplify it. There probably be some YAML somewhere, but it hopefully it'll be straightforward. It's just a really nice.

You know the whole industry has gotten used to this deployment mob. We use containers. I don't. You know everyone I know uses containers for for a lot, of, a lot of different purposes in use cases. So like It's really, it's just really nice Logical evolution and it feels like a very clean way. Obviously there'll be trade-offs. Obviously there's gonna be gotchas there always is with any platform, but it does feel like a very exciting thing. And then, by the way, that the way they demoed it was a very good way to show That excitement they had. They had like what? 10 10 people standing behind the state. They did a 10 people demoing at once, like running different apps, and they had like a 10 screen kind of zoom screen with 10 10 video streams of Different screens, different laptops showing 10 different applications, all very different to each other. Yeah, so so you know, super excited to explore it, particularly for us, for propel. I think there's an obvious fit there, but lots to learn and looking to.

0:30:32 - Tyler Wells
I Forgot how bananas that demo was where, you know. You know I think you had the, the one dev advocate, up there. You know that had been doing all the stuff that was, you know, wearing the snow goggles and everything the whole time and it was like the screens behind them were all perfectly opaque. You couldn't see anything. Then, all of a sudden they went translucent to a degree and there were all these people behind there Running demos and then they started showing them all on the screen and that was. That was pretty incredible, yeah the the.

0:30:58 - David Ganly
They started with hacks. They showed hacks on as running entirely containerized, including the whole kind of notebook style UI. I think they showed it like connected, running on, running with a notebook, with them Nvidia GPUs as well, so demonstrating again the connectivity. You know the partnership that they've just done. But you know that was whiz, bang enough, right, like, wow, they're running the whole of hacks inside a container, inside the snowflake platform. But then, yeah, they just kind of up the ante one more time with that. Let's, what was it? Let's not just show one more demo. How many do you want? and then they did.

0:31:30 - Tyler Wells
Let's do 12, let's do a thousand. No, they did. Often they all worked, so it's very cool. All right, let's go prediction time and first prediction I'll go. Niko Does propel have a booth next year at summit 24?.

0:31:47 - Nico Acosta
That's a good one. I would, i would. I would find true to say that we, we do. I wouldn't predict the size of the booth Yeah, the biggest one, the biggest one, the biggest one But I would say that we do. I think the where we see our platform going, where we see Snowflake going with container services, where we see the community going, is just incredibly well aligned And I think it makes sense for us to start having a bigger presence there.

0:32:34 - Tyler Wells
I mean it was a great ecosystem to interact with. It was a great ecosystem to talk about a number of the problems that we help our customers solve. It was a great ecosystem to learn.

0:32:47 - Nico Acosta
And I saw that next year they're going back to San Francisco, So they're doing Moscone.

0:32:54 - Tyler Wells
Oh, wow, okay. That'll be that'll be very different. That would be a very different. Yeah, a very different type of yeah, going from, like, all the chaos that is Vegas to San Francisco Moscone. That's a very different atmosphere. Yeah, very different atmosphere. Probably less hangovers, i bet, in the mornings for many people.

0:33:18 - David Ganly
But then again I don't know.

0:33:19 - Tyler Wells
There was definitely some folks that were hurting. I think it was on Wednesday. Yeah, especially the developer one, dave. What about you Predictions?

0:33:29 - David Ganly
Oh, just the one of the things we talked about previously before this was the, the kind of the app marketplace aspect. We talked about container services, but the actual kind of the you know, having that app marketplace much more built into the snowflake experience. It's going to change how a lot of those data B2B companies approach distribution. If you're inside the snowflake ecosystem, you can kind of be part of that snowflake themselves on the stage. We're talking about like monetization right, they have monetization tools built into the kind of the app platform now. So, yeah, predictions you're going to see, you know.

You know we talked in the past. We talked about like when did Salesforce get its first company built on top of Salesforce that hit IPO? right, it feels like that's going to be what we'll see with snowflake in the future is startups that are started and exist only inside this marketplace, inside snowflake, becoming million dollar and then billion dollar companies in themselves. And you made the joke earlier about SOC2 and compliance. Yeah, Like you start a startup inside this ecosystem, you know you immediately get that distribution, you immediately don't have to think as hard or you know in depth about compliance. You immediately have a different way. So I would be surprised that we don't see and I think they already showed. You know the earliest adopter. They had a guy on stage from a startup who was like only available in the marketplace.

So I said it's a no brainer that we're going to see more of that And I think we're going to see highly successful companies in that kind of change how people build their companies.

0:35:03 - Tyler Wells
Yeah, the advent of the new apps that are going to be built in those containers, going to be the native apps that are going to be able to bring all of this you know, power of generative AI and everything else across all of your data, is going to be pretty interesting. And then tying that all into streamlet, then tying it into you know the other possibilities of visualization tools, And then the applications that now become part of the customer facing fabric, I think is going to be super exciting and watch how that grows.

0:35:32 - Nico Acosta
Yeah, you know, you know when, when a marketplace like that works, when you have a solo founder that became a millionaire with their app on the marketplace.

0:35:47 - David Ganly
It's not that dissimilar. I mean it's a different space, but it's the paradigm of, like the app stores, you know, like the iPhone, like it fits there, where you'll see these early adopters who get traction really quickly And then over time you're going to see the resophisticated players getting involved. Probably a lot of the I'm assuming, a lot of the financial yeah, we saw a lot of financial institutions there, either with booths or just talking about their you know snowflake installations. So just within that one vertical of financial institutions sharing data, building these native apps, you're going to see this explosion. I assume. I'm assuming stuff like already has people knocking at the door to get in. Get into all these previews.

0:36:32 - Tyler Wells
I think there's one final thing to discuss, and it was the picture, dave, that you shared on the final evening of that massive vegan burger that you tracked down.

0:36:41 - David Ganly
Oh yeah.

0:36:42 - Nico Acosta
Yeah.

0:36:43 - Tyler Wells
That was incredible.

0:36:44 - David Ganly
Yeah, so.

so I think you know Vegas is famous for having really good food and there's loads of really good restaurants, but I think most of it is mostly catered to the kind of more mainstream. So you know Gordon Ramsay, but all these delicious things, and so I set off on a quest to find a amazing vegan restaurant which took me away from the strip, which took me in an Uber 15 minutes away into Northern LA, but I did end up with a, possibly the largest the photo did not do it justice like maybe as larger. I have a large head, i think, but larger than my head, which no one believed me was vegan, which is also a win, right, because you literally thought that looked like a very good bacon cheeseburger with like three or four patties in it, but actually it was completely vegan.

0:37:31 - Tyler Wells
It was massive And from the picture when I saw it the next, I was like you could not tell. It was near impossible to tell that that was vegan. It had all the fixings. It looked 100% legit And I would eat it. I would, I would, I would go out on a limb and I would give it a try. I think it's pretty good.

0:37:50 - David Ganly
If you're watching, I do not remember the name of the restaurant, which is a terrible faux pas for this podcast. If you're watching or listening to this and you want to like, drop a comment on the LinkedIn and I will send you. I will drop the actual restaurant name because I feel really guilty that I went in pursuit of this thing. I found it and now I can't remember the name of the restaurant. You caught me off guard with that amazing last question.

0:38:12 - Tyler Wells
Yeah, that's the point of the podcast. right Of being the host, I've got to do that every once in a while. Throw your curveball, make it, make it, you know, make it tough for you.

0:38:20 - David Ganly
I actually had to get fun 95% data, 5%. where does Dave get his vegan burgers? I get it.

0:38:28 - Tyler Wells
Vegan burgers from, it's perfect, and on that note we should probably wrap this up. Otherwise we'll I don't know maybe get into what was our favorite steak, but you know there's so many.

0:38:38 - David Ganly
You guys, you guys pretty much had some pretty amazing steaks. I was party to that, so yeah, some good big steaks.

0:38:46 - Tyler Wells
Yeah, the food in Vegas is always amazing and there's just so much of it and it's easy to walk right in. I I fear, like when it's at Moscone next year, you're going to have to get reservations way in advance. There's no way to get the good stuff. You'll be, you'll be kind of out of luck So well, perfect, gentlemen.

That was a lot of fun. I'm very happy that we all got to go to the summit this year and got to interact with our customers and potential customers and future customers and other folks that want to build amazing apps on top of their snowflake data and look forward to what we can do for them in the future, specifically on snow park containers. I think that's going to be a lot of fun and look forward to it. So thanks for joining me on this version, or this episode of the data chaos, and we'll talk again soon.

Transcribed by https://podium.page