Snowflake Inc. (SNOW) Earnings Call Transcript & Summary

Hi, everyone, a warm welcome to SPN Pulse. Let me start over, right? I forgot to unmute myself today. I'm Hwee Bee Tan, I'm the Head of Partner Marketing for APJ at Snowflake and Partners, thanks for your time. It's been an incredible week following our summit in San Francisco, where we welcomed over 20,000 attendees doubling our growth this year. We're excited to bring that momentum back and continue our journey with all of you today. Today's session is all about what's new, what's next and how we can continue to grow and win together this year. You'll hear directly from our leadership, Mike Gannon, our CRO from Snowflake, and learn from the inspiring journey from our customer, Spark New Zealand as well as RelationalAI, right, how they're transforming and leading the way with us. Today, we also have our latest product innovations that is brought to you at Summit. We're having Amanda Kelly to share that with you later. So stay tuned to the very end of this session. To kick off all the things, I'm delighted to hand over to Ash Willis, our VP of Partner and Alliances for APJ today. Over to you, Ash.

Great. Thanks, Hwee Bee. Good to see everyone. Hey, Hwee Bee what are -- what shirts are you wearing? Is that the L.A. Olympics, right? Good stuff. So I missed out on grabbing one of those. But for everyone's benefit, we announced last week that Snowflake were a sponsor for the LA28 Olympics, which is super exciting. And I'm waiting for the announcement around Brisbane '32 as well just to get ahead of things. So good stuff. Thanks, Hwee Bee. And Mike, welcome to our APJ SPN Pulse. Good to have you here.

It's great to be here. Thank you for having me. Exciting.

And I know you're joining from the East Coast of the U.S., so quite late, and we appreciate you jumping on board.

Glad to do it. I'm getting used to these late nights to talk to all my Asia Pacific leadership team and customers. So glad to be here.

The joys of being the CRO of an international company. We're super excited to have you on board, Mike. Just over 3 months at Snowflake now. I know you've been super busy with a whole bunch of activity. I do want to talk about last week at Summit in a moment. But let's kick off with our latest earnings announcement. So great time to come into Snowflake.

I know you had the opportunity to do the earnings call from the New York Stock Exchange. So maybe share a couple of observations in terms of Q1 results. A quick intro to yourself, your background and why you chose to come to Snowflake.

Yes, sure. So for those that have not looked at my LinkedIn profile or had the chance to look at me virtually. So I joined Snowflake 90 days ago as Chief Revenue Officer. I've spent the better part of the last 29 years, really in 2 -- really 2 companies. The first 15 years of my career was with EMC Corporation. And I started my career just as the third industrial revolution was kind of taking shape in the Internet, for the most part, was changing the way that we are operating business to business as well as business consumer. So I had a lot of great experience and fun driving transformation and building out infrastructure for companies that were starting to deal with more e-commerce applications and transforming the way they operated from a technology perspective. And I then took a shift over to VMware, and I just wrapped up the General Manager of the Americas at VMware and I went over there because I started to see that software was starting to eat the world and people were delivering more value through software-defined data centers. So obviously, as the leader of virtualization, we started to draw a much greater expansion of virtualization beyond compute into storage and networking and really helping customers build private clouds. And inevitably, we built out partnerships with all the major hyperscalers because we recognize customers want to move their applications to the hyperscalers. So we created this. VMware private cloud in each one of these hypers is giving customers the ability to vMotion applications between on-prem and the hyperscalers. And certainly got a lot of experience watching customers trying to modernize their applications to move it to modern cloud. And I think that was certainly a very difficult challenge. But after 12 years of VMware, it became apparent to me that customers started spending more time and effort, modernizing their data because they recognize they're going to get much more faster path to value by driving a new data architecture as opposed to trying to lift the legacy application and moving it to the cloud, which wound up being much more difficult than people had ever anticipated. So as I start to look at really the next 10, 15 years of my career, I knew I had to get into the data space. I was afforded the opportunity to interview for this role, which fortunately enough, I was able to get. So I'm 90 days into the role. We kicked off the year with some tremendous earnings. We had our first $1 billion revenue quarter, which represented 26% growth year-on-year, which was a fabulous way to start the year. Our net revenue retention is 124%, and that metric basically suggests that customers are not just doing a static renewal. They're expanding their contracts with Snowflake. So demand continues to expand with all of our customers. What's resonating well is I'd say our remaining revenue obligation, which represents really the backlog is $6.7 billion, and that's a 34% year-on-year increase. So customers are signing up for longer-term contracts. That backlog obviously is an addressable market for everyone on this phone call to go help figure out how we can drive migrations, analytics, AI, that's really the attack surface that we've left open for our partner community to come in and help customers drive consumption and get value out of it. So our remaining revenue obligation continues to be very strong. And of the 11,200 customers that we have total, we added about 451 in Q1. And approximately half of those customers today, so about 5,200 customers, so about half are actively using our AI and ML products. And that to me is a really strong leading indicator of how customers are looking at Snowflake, not just from a data perspective, but really trying to simplify their AI initiatives. So we're seeing great adoption around the AI. So great way to kick off the year, a great way for me to learn what customers love about VMware. And again, of all the customer meetings I've been in, I consistently hear we're easy, we're trusted and we have this great connected capability of allowing zero-copy data shares between 2 different Snowflake instances. So we're starting to create this market network effect around Snowflake. So really just great, great opportunity for me to see the company on display in Q1, and see some fantastic results. But I think, as I've always said, the market votes with their wallets every day. And right now, they're voting on Snowflake and it's an exciting time to be here.

Yes. It's an incredibly exciting time to join, Mike. I think anytime a company sort of celebrates their first $1 billion quarter is amazing. But as you touched on, this momentum that's building around data, and having been at Snowflake now for just some 3 years, when I first joined, Snowflake was doing some amazing stuff but it was very much pitched kind of in this technology space. But I think this trend around sort of driving business impact and the value that we can add to business is really impressive and really becoming very front of mind. So one thing I just want to unpack a little bit more that you spoke about there. So $6.7 billion RPO, maybe just dig into that a little bit more and what that means in terms of opportunity for partners because I harp on this all the time. But I think it's good for this audience to hear from you in terms of what that opportunity actually looks like?

Yes. So we have a unique compensation structure at Snowflake. We don't get paid when we book a contract. Our sales teams only get paid when we drive consumption. So that means, we've convinced the customer that Snowflake is an essential data and AI platform. They buy into a multiyear contract with us, which is what we're seeing trending wise, as customers are moving to multiyear contracts. And then we start looking for use cases, what data sets are we going to help them modernize. The low-hanging fruit from us continues to be legacy traditional warehousing technologies like Teradata, Netezza, Oracle Exadata, SAP HANA, legacy warehousing that frankly isn't scaling to the needs of the company. So we're doing a lot to drive migration. So there's an unbelievable opportunity for us to partner with this ecosystem to help drive and accelerate migrations onto the platform. So when I say that we've got this large backlog pending, you're going to find a very receptive audience at Snowflake that wants to partner with organizations that can help accelerate migrations onto the platform because that's going to drive consumption, and that's when our sales team gets paid. So it's a very unique structure from that perspective. But the thing that's really landing on me when we talk about AI, most customers look at us bringing a legacy transactional, very structured data set into Snowflake. And I'll give you a great use case example of where we did this for a very large construction supplier in the U.S. We had already brought a very structured data set into Snowflake, and we're helping them running some advanced analytics around typical structured data. But one of the things that they had was -- the challenge was they've got several people that run -- their task is they build bids. So people will send them a blueprint, an architectural blueprint and you will have a person accept that blueprint. And over the next 3 days, they will count how many windows and doors and trusses and 2x4s and they build out a bill of materials and they submit their bid, and they have about a 20% success rate where people will contract them for supplying the builder for that architectural blueprint. Well, we were able to showcase in a presales motion when we were able to bring that unstructured data that blueprint into the Snowflake environment and map it up to their structured data, we were able to basically use an AI model to go through and analyze 100 blueprints per day as opposed to 1 every 3 days. So when a customer sees the unlocking that value and now that person can submit 300 bids a day versus 1 over 3 days just by bringing an unstructured data set and a structured data set together and using a basic AI SQL query, it became a very powerful unlocking mechanism. So what I think most customers are really starting to recognize is when they bring these structured and unstructured data sets together, they get a much richer look at their data ecosystem and they're leveraging us to really simplify the business outcome. So that to me was really an unlocking moment where the customer wasn't sure we could do it. So most customers lack the awareness that this -- the powerful platform we can deliver. So that to me, became an eye-opening moment for me to say, this is where customers are struggling as they don't think this stuff is capable, but we can really deliver some pretty significant outcomes just if we ask the right questions. And I think what I'm asking my team to do is let's assess the business outcome first and then figure out how the technology will deliver the outcome. So that to me is a lot of -- that's what I'm working on with my team is making sure we identify the business opportunities and the business outcomes. The technology is the easy part, really.

Yes. I think it's a trend that we see out here as well, helping sort of customers imagine the art of the possible. Everyone wants to talk about AI. Everyone kind of recognizes that technology can add a bunch of value to their business. We're really trying to understand what those key use cases are. And I love that one that you just described because not only is it improving productivity, but I would imagine that it's more accurate than somebody trying to count out windows by hand and all that sort of stuff as well.

Yes. But I think to your point, it's not a job elimination. It's not AI taking over the role. It's making that 1 person far more productive. And I think that's really the unlocking moment right now as everyone scared AI is replacing jobs. And listen, eventually, it may. But the first iterations of what we're seeing is we're really just unlocking productivity and allowing customers to accelerate revenue streams and potential. So the first unlocking moment, I will say, is really leveraging AI to drive productivity. And I think what's going to happen is you're going to see a skills shift, right? So what might be a radiologist is doing today, maybe we don't need to hire 30 new radiologists in the health care market, I can scale my business more with the existing head count I have. And that to me is, I think, the power of AI, at least within its first iteration.

I was chatting to a radiologist at a forum earlier this week, actually, and they were saying, always, hey, what's AI doing? And they were talking about in terms of improving accuracy and to diagnosis, not reducing the number of roles but providing better outcomes at the end of the day. Mike, I'm conscious of time, and I've got so many more questions.

Let's get a couple of more to speed round.

Let's switch to Summit. So last week, we were joined by more than 20,000 of our great friends in San Francisco, what were some of your initial thoughts other than [indiscernible].

It was eye-opening for me on a few -- again, I'd shared I just left VMware after 12 years. And I think VMware started their user conferences in about 2003. And at our peak, just before the Broadcom acquisition, we had about 18,000 people there. And to see Snowflake only after the sixth year of having our conference, right, we doubled our capacity from last year at 11,000 to 20,000 this year. So unbelievable amount of people. I was not expecting that big of a show. So I was a bit blown away by that. But I'd say the quality of people that we're showing up to the event was really eye-opening to me. I mean we had Chief Data Officers, we had Presidents of AI. We had people that were truly trying to lean in and understand how Snowflake can unlock value for them and deliver better outcomes. So obviously, we had a large degree of analysts and architects and AI engineers. But I also saw a pretty good concentration of business minds coming to Snowflake and trying to look for the art of the possible talking to other customers about what they were doing with Snowflake. So there was an unbelievable network effect of customers learning from customers. And what's really interesting is 70% of the content that was delivered at Summit last week was delivered by our customers, right? So it's one thing to hear a very biased opinion coming from us. But when you hear how customers are leveraging the technology, that to me was an incredibly powerful statement. So I was super impressed by the content. I thought the logistics was fabulous. The energy level was high, but to see how our customers were contributing to the Summit, I thought was just awesome.

Yes. So a couple of customer highlights like seeing [ Camber ], like a very big customer that we're incredibly proud of out here in APJ to be featured as part of a keynote was amazing. But I'm not sure if you're aware, Mike, like we had so many submissions from customers and partners that wanted to join us to advocate for Snowflake at Summit to present that we could only take such a small portion of those submissions. So...

It's an unfortunate byproduct of having a popular show as you can't have everybody. There's only so many hours and sessions we can hold. But I do think next year, we're going to have a bit of a problem. If we get to 30,000 people, the Moscone Center is going to get really crowded. So -- but we are excited to take our world tour. So for those that could not travel in to San Francisco, we've got a world tour going all around the world. So we're going to bring Summit to you I'm going to try and hit a few of them as well. So we're excited to bring the content for you to Asia for those Asia Pacific that can travel in. We're coming to a market near you soon. So look at our registration sites and get yourself registered. We're going to have some great events on the road.

Yes. We're excited to welcome [indiscernible] for a few of those events. Last point on Summit, we were chatting at the start of this call in the green room, and you were saying you had an opportunity to spend some time with the Snowflake investor community. I think there's always some really interesting insights that come out of those conversations. Are there any kind of key things that you could share with this audience?

They had quite a bit of -- they had a lot of questions of -- so I was on the panel. So we had 200 investors in the room. It was myself, Sridhar and Christian, our Head of Products. There was a lot of questions that were certainly around product and strategy in the future. But the questions I was receiving was really relative to, Mike, how are you looking at the go-to-market function? And as I explained to the investor community, there's really 2 things I'm focused on, which is, one, is how do we scale our business efficiently. And number two, is velocity. How do I remove friction from our selling motion? And what I shared with the investor community was my desire, a very strong desire to build an incredibly strong partner ecosystem, right? So our global alliance and channel is going to be a big part of our future go-to-market motion. In order for me to scale this business, I can't keep hiring direct sales reps. We're going to be making a major investment in our channel and distribution. So we've got wonderful relationships with the hyperscalers. There's some emerging needs for sovereign clouds in Europe. So we have to make some investments and some unique investments in Europe. And there's probably some of that even in Asia Pacific, where we have to build sovereign clouds. We talk a lot about the ISV marketplace, right? These are people that are building businesses on top of Snowflake. It's an incredibly rich channel for us as well. And then we've got incredibly strong systems integrator market where companies are leaning on systems integrators to help them drive modern data architectures and platforms. But the fourth and what I'd say, frankly, is an underinvested leg of our channel is our distribution network with traditional resellers. So I'm spending a lot of time what I'd say, ramping or rebuilding out what I think is going to be a very rich partner resell program. So coming soon, you're going to see a big investment from us in resell. We're excited to launch this. This, to me, is coming from EMC, coming from VMware. I saw the power of the channel. I believe in the channel. And I'm going to be making some big investments there. So that was a really big part of my presentation to the analyst community is how we're going to scale this business. We're doing quite well already. We're going to grow the business 26% as we've mentioned to Wall Street. We raised our guidance at 26%. We should be doing well north of 35% if I activate this channel. I shouldn't say if, I'll say when I activate the channel.

So the investors, obviously, like that message.

The stock went up that day. It was good.

Stock went up and ratings went up as well. But Mike, just to close out, in the last 3 months, you've already made some pretty decisive moves to make us a more partner-friendly organizations and things around comp neutrality for marketplace. And as you said, there's a bunch of other really exciting stuff to come. I could keep asking questions for a long time, but Hwee Bee is going to boot us out. So thank you so much for joining us. Thanks for investing in this ecosystem. And we hope to see you out here very soon and get you back on Pulse, maybe in 6 months' time and get a few more updated reflections once you've had a bit more time in roll as well.

Yes, I would love to do it. Thank you so much. It's an honor and privilege to be on the show with you all tonight. I look forward to getting into the market soon and making my trip through Asia Pac and seeing as many of you as possible face-to-face. So thank you so much. Pleasure meeting you all virtually.

We're a great host, Mike. So you have a good time.

Very good.

See you later. Thanks.

All right, take care. Bye. Bye.

Hwee Bee, you're on mute, if you're trying to do an introduction.

No, I'm just saying a good one for Mike, and we have our customer and partner Anshuman and Marco is here.

Good stuff. Hey, guys, thanks for joining us.

Thank you for having us.

Great to be here.

Good to see you both. And I know that there's some really good work going on between Snowflake, Spark New Zealand and RelationalAI. But before we jump into that, Anshuman, why don't you give the audience a quick introduction to Spark New Zealand. It's obviously a brand and a company that's very familiar to me, but maybe our -- not all of our audience are familiar with you guys.

Certainly Ash. So Spark New Zealand is the largest telco in New Zealand. New Zealand is not a huge country, but...

It is a little country just off the side of Australia, right?

It is, right? You can miss it easily, but we should...

Beyond the -- now I'm going to get a lot of group about that statement from my Kiwi colleagues.

Yes. So yes, the largest telco, we have the highest share on mobile in New Zealand, but we are also significantly into the business and small and medium enterprises in NZ. Small and medium enterprises make up 97% of New Zealand's businesses, and we play a significant role in there. And 45% approximately of our revenue comes from enterprise and small and medium enterprises. We are an agile organization overall. And we went agile quite some time back about 10 years back as an entire organization. And we are really, really agile in the sense that we are very open to using technology, trying out things, experimenting and interested to innovate and try new things out. So I'll probably stop there.

It's good. That's a super helpful overview. Marco over to you may tell us a little bit about RelationalAI and your role.

Yes. Thank you, and hi, everybody. So look, RelationalAI is a technology company headquartered originally actually in Palo Alto, and we've been partnering with Snowflake since very, very early on because there are a lot of actually ex-Snowflakers part of the team. The goal of the company is really, in my opinion, very simple, is to enable organizations, partners and customers of Snowflake to empowering them to make better decisions on Snowflake, okay? And there is a true -- it is an inner belief that organizations pretty much live and die by the quality of the decisions that they make. And what we want to do is to make the process of making decisions within Snowflake across a variety of different use cases and questions as easy as possible to the end user. We are approximately 200 people globally. So still quite small but growing. We are fully remote, and so we are spread across the globe, I think 50-50 between the U.S. and the rest of the world. And my role is really to lead the go-to market team globally and that really encompass like sales, delivery, marketing and partnership. And so it's a real pleasure to be.

Good. And we've -- as you said, we've got some really good work going on together. So good stuff. Anshuman, like, let's -- you mentioned Spark is an agile company that's been my experience working with you guys as well. But I know that you guys also have a strong vision for AI and leveraging AI is kind of a strategic lever to drive business growth. Could you tell us a little bit more about that and what that looks like?

Okay. I think when most organizations start doing AI or start getting into AI, it is generally for incremental benefits. The way we look at it is more at an organizational level, right? Can you look at the potential of AI to change the way the organization works? Am I reimagining the organization in a world with AI? Am I taking the current processes incrementally or taking a more holistic approach. You're not building for the future. Now when we build our architecture and we build for our use cases, we are generally taking a lens, which is a longer-term lens. A problem statement, can we take the problem statement? Let's say an example of it is provisioning a complex product, right? And there are multiple steps to it. That problem statement might come to us from a provisioning team. but we extend it out to a bit further out to say, at some point, can we enable customers, right, to come in via self-service and actually provision a complex product themselves, right, which is extending it beyond the immediate customer who's asking for it and taking it to a level where AI can go and make it really, really streamlined for customers. Now there might be multiple steps to that journey, right? Step 1 might be that we build it for provisioning because the stakeholders are from provisioning team. Step 2 might be taking that same thing and making it easier for our sales teams to sell the product. And step 3 might be actual customer enablement and self-service of that particular outcome. But when we build the architecture and we build that use case, we take a lens as to where can AI go, right? What's the potential of this use case and we build for that use case so we can extend it over a period of time in the next 6, 8, 10, 12 months. At the same time, the other thing that we also consider that to be able to do it, we need to have skills of the future. And as we can see what's happening in the world of AI is a lot of skills, which are about coding, right, a lot of skills which are about functional knowledge on a particular topic. They are becoming more transient. You can acquire those skills easily. Core skills are tending to go more towards creativity, problem solving, et cetera, right? Are we bringing in those people into Spark, the best of those people in with us on this journey? And at the same time, are we also providing training to everyone else in the organization in terms of how to use AI, how to survive in the world of AI? Now, we are on a journey on all of this. We are not the best or the perfect in this, but that's the lens that we take when we look at the potential in a world which is where AI will play a significant role in the life of each one of us going forward.

Okay. I'm going to jump around a little bit because I think that you're touching on a few interesting topics there, but can you share maybe 1 or 2 examples of where Spark NZ have already adopted or already built certain AI-powered tools or initiatives?

So I think one of the examples which I'm really proud of is something which is about call summarization. And I'll give you the context of this problem statement, which is, we get x million calls in our contact centers. And every call used to require an agent to spend 2 minutes to summarize the call at the end of the call before they log it into CRM.

You mean a real agent in this instance?

Yes. Now it's to classify very carefully, like when you're talking about humans versus robots, robotic agents. So yes -- and I think taking the call is important because that's the value-add providing customer experience is important. But logging a call, right, with human -- with your interpretation of it is not the most value-added thing, and then we had an industry of compliance around it, right, to say, our agents doing it, et cetera. So what we did was we brought AI in the life part of the call, right, where as soon as the call ends, AI summarizes the call. And my personal idea was summarize the call and log it, but we have to work with stakeholders in all of these areas. So our ask from contact centers was you cannot log it, you have to play it back to the agents for them to validate if it's correct or not, right? And we had to play it back on a live call path and they update it and then log it into CRM. What we observe is the number of times people actually updated is nearly 0. And I'm waiting for a day when we'll flick the switch and will get automatically logged into CRM. But that's a real example, live example of AI in a live call part, which we have implemented, which is generating business value. The other impact, benefit of such things, which you don't anticipate is that the quality of data that we are generating about each of these calls is much -- of a much higher quality compared to what was getting logged earlier, which is downstream benefits in terms of root cause analysis and then reducing the actual calls or issues that customers face.

That's a really interesting example. And it sounds like it's returning some pretty impressive business results. Marco, talk to us a little bit because I know you have a great partnership with the team at Spark. And talk to us a little bit about the work that you're doing there, and I know there's some stuff around relational knowledge graphs that are involved in this project as well. Maybe you could share a little bit of context.

Yes, absolutely. And I'll start maybe by briefly telling you the people here, while we think it's really, really crucial actually this concept of capturing knowledge, which really links to what Anshuman was saying here, capturing really data and knowledge. And look at the core of it, again, there is a belief that every organization is different and knowledge is captured in a variety of different ways within organizations, right? You find knowledge across spreadsheets, you find knowledge into our databases, you find knowledge into application code, you find knowledge in documents, you find knowledge in people's head.

And structured and unstructured.

And structured and unstructured. And as we said like in [indiscernible] like knowledge is spread really, really everywhere. And that knowledge fragmentation generates a lot of pain actually because it's really difficult to provide a unified a cohesive view of your enterprise. And so sadly, when somebody asked a very simple question what are my -- what were my sales last week? Now teams crumble, right? Because they're like, okay, it's a week 5 days, Monday to Friday? Is it a week 6 days? And do we include all the products here? Is this flag should be true, should be false? How should we interpret that question. And by the way, that's a problem that is true for people within the organization, but it's also true for the newcomers into the organization, right? They need to learn how the organization operates. And in the era of agents, agents are little -- I call the agents, little interns, little employees in your organization. It's true also for agents, right? They need to learn. They need to understand how enterprises operate. And so really, the point of a relational knowledge graph is to solve this challenge of this knowledge silos. And it's interesting back in the days, remember, Snowflake we said like, hey, our common enemy are the data silos. I think now I make the statement, our joint common enemy beyond the data silos are the knowledge, the knowledge silos, right? And so what that does then, it enables organization to describe to store, to share, to discover and then infer a new knowledge. So now why is that important, right? And maybe -- and this now links to some of the work we're starting to do with -- of course, with our friends at Spark here. Well, it's -- in just a few -- just to give you an example here, right, in just a few weeks, the team from Anshuman here leverage RelationalAI knowledge graph to build a reliable digital twin of their network to simulate then complex users and network behaviors, okay? And that now implies that actually you get LLMs that now can become creatively smart, okay? They can be actually fine-tuned to their knowledge graph, so they can understand the business domain. And then thanks to a set of what we call reasoners in Snowflake, fully embedded within Snowflake by the SQL reasoner, a predictive reasoner to reason about the future, a prescriptive reasoner to reason about what to do about the future. Now this system can really answer all sorts of questions, okay? They can answer questions about the past. They can ask -- they answer questions about the present, the future and then what to do about it. And so that is why such a technology and reasoners together are becoming important and really, really central to, I think every organization will become more and more important moving forward.

Yes. It's an incredibly interesting space and like this idea of knowledge-centric and really understanding those different perspectives, yield some pretty powerful results. Marco, just a little bit around like how do you guys integrate with Snowflake to help make some of these outcomes a reality for customers such as Spark NZ.

Yes. Look, again, the belief here is that we need to meet the users like end organizations where they are, right, and really be able to complement their past, existing and future investments, right? And so where users are nowadays and where enterprises are, they're on Snowflake. They are actually storing all sorts of data, on relational systems, right? And so it is important to be able to then provide support for users and enterprise within those relational system to build and deploy AI at scale and being able to actually build intelligent applications. And so what I'm talking here is really being able together, like with Snowflake and with partners like ourselves or with more partners, thinking through what I call a unified relational AI infrastructure. And apologies here for the play in words, but you see what I mean when I say relation because it's built on a relational paradigm. That has a set of key characteristics really as a data centricity, so you get all the data types being able at your disposal, right? You get as you said, actually get structured data, you get structured data, you get images, right? You get documents, you get graphs, you get tables, but it's all there for you. Then you get a set of reasoners available for you, again, baked in into the infrastructure that helps you to actually reason on that data. And get your secret reasoners, you get your vector search reasoner, you get your graph reasoner, you get your predictive reasoner, prescriptive reasoner. But these are ways for you to interact with your data. And then you get a semantic layer, okay, on top of it, which is, in our case, is a relational knowledge graph that enables you to model the business and then express rich semantics that go beyond the type of things you would do with BI or dimensional semantics. And then last but not least, because nowadays everybody is talking about the ability to just question -- ask a questions to your data, a post trained large language model that is really fine-tuned to that domain so that now it understands how the business operates and can use those reasoners to ask questions to the data. And this is really a paradigm shift as you called it out, Ash, because in the past, we will be pretty much application-centric, right? We will take data and ship data to the application. That's what we'll be doing. And what we are seeing here, if you have now a unified AI infrastructure, you are much more knowledge in data centric, okay, where you are actually sending -- not sending, but really bringing compute to the data and the infrastructure is now providing you with the capabilities that historically you would either buy via point solutions or you need to build and maintain yourself. And then this approach really dramatically simplifies infrastructure, accelerate innovation and unlocks the full value of really data and knowledge together.

That's a really comprehensive answer, Marco. But I think that you touched on a key point there right at the end, which is -- there are lots of great points, by the way. But you spoke about accelerating innovation. So Anshuman, to close out, how has this partnership between Spark, RelationalAI and Snowflake helped you guys accelerate innovation?

I think I'll just probably touch upon a few of the points which Marco mentioned, which I really like. And it's the fact that at the end of it, I mean, for any AI or any of the things that we are doing, data is the oil, and whatever AI we are doing on top of it, that is, again, creating more data and more information. And for us, moving into the journey of moving our data into the cloud, right, with Snowflake was the starting point of a lot of the AI and innovation that we have been able to do. We started off with the general idea of moving data into the cloud from why people move into it from a cost, scalability, et cetera, perspective. But that has kind of put us on the path of this journey on exploring and experimenting with this data and creating new things. The things which Snowflake is bringing to the table in the form of partners and partnerships is really, really interesting and important for us because I had a feeling that the world of the future will be a world of partnerships because there's so much innovation, which is possible now because -- and again, AI is making innovation possible to quite an extent. And I see us having a lot more partnerships going forward than we have now. And the ability to onboard such partners, right, using the Snowflake AI and data cloud easily, seamlessly and bring them where the data is and then use that in a meaningful way, I think, is of immense value to us. The other thing I wanted to also mention is I think we've gotten an awesome account team from Snowflake, which works with us because we are very demanding sometimes, right? We ask for -- we have asked for like the Cortex functionality 3 years back because we thought this was possible. And I keep pushing Richard and the team to put us in touch with your product team and partners, and they're extremely helpful, and they bring the teams in which I really, really appreciate as well, and it's a great working relationship for us. So I see it's a very strong relationship, and I see it growing with new partners coming in, and the ability to work with partners like RelationalAI is great for us and my teams.

That's good. We talk a lot about accelerating time to value. And I think hearing those stories is really valuable. And I'm so pleased to hear that you pushed Richard and the team hard down there and the results are definitely showing. So we're at time, Hwee Bee is going to boot us out in a moment, but Anshuman, Marco, thank you both for joining SPN Pulse. Lot of great insights and really enjoyed it. There we go, some photos from last week.

It's great.

Amazing. I'm the most handsome in this group. No doubt.

Well, it's good to see some of those shots and it looks like you got to do a couple of videos there as well. Fantastic. Good stuff. Thanks, folks, and over to you, [ Chris ].

Amanda. Amanda will be joining us. Yes. See, I think, Amanda, you need to click the button on the right hand, Yes. Yes. Cool. Hi, Amanda, how are you?

Hello. I'm doing well. Excited to be here.

Okay. Awesome. So Marco and Anshuman, you can, no, on the back -- you can enter the backstage, and then we can have Amanda share with us. Over to you Amanda to share with us the latest innovation of Summit.

Yes. I'm so excited to be here. I was actually just in Korea and Japan a few weeks ago. I know I need to make it out of Singapore, too. But then I was at Snowflake Summit, we announced so many amazing things for Snowflake and for you as partners and customers. And so I'm really excited to talk to you about really the future of AI-driven data platforms, right, which is what we're building at Snowflake. And we truly are evolving from a data cloud to a fully integrated AI application platform that's redefining, right, how products, applications and intelligent systems are being built, right? And we're giving you just the download today of some of the top, top things, but really, you should go, you should watch a bunch of the videos online, so many cool announcements, so many new quick starts, things in our solutions center that you can get started with and try today. So I wanted to talk to you just a little bit about our mission, just kind of fit in where we're investing and how these things are working, right? Obviously, we talked about we're becoming the AI data cloud. And we are still focused on giving you that 1 unified managed platform that's going to make it easy, right, for you to connect, bring all your data together to trust it, right, to make sure it's governed it's secure and then to do some pretty cool things with it, which I'll show you about in a second. So next slide. This is kind of a simplified view of a lot of the things that we're investing in, especially on the data side, right? You've got all these different data sources. Many of them you've already brought together in Snowflake, but there's probably a lot more that you want as well. right? I often think about as a product lead, I often only see a slice, right, of what we're doing. And it's frustrating because you don't have the full picture of what your customers are doing, what's happening in the market, where you can only get that by bringing in more data into more applications that help you join these views together, right? And then, of course, right, you have -- the hard part. You've got to get all that data process. You got to get it into a place that you can actually do things with. And then once you have that there, right, we can use AI in so many interesting ways on the BI layer for data science, for apps, for sharing that makes it really easy for you to actually drive decisions and outcomes in your company. Next slide. So as we look at kind of the overall Snowflake platform architecture, right? It always starts at the bottom, right? It starts on the cross-cloud, right? No matter what cloud you're in, what region, right, we are able to support you. We're able -- that's why we're able to support so many of the largest companies, right, in the world globally, right? We are a unified AI-ready data platform that starts at the data layer, right, unstructured data, semi-structured, structured, it doesn't matter, right, what data type it is. It doesn't matter what format it is, iceberg, right? Is it a hybrid? Is it OLTP, right? We're going to support it, right? And then we have data lakehouse, we have data warehouse. We have ways to work with process that then to interact with it, right, to transform is it's equal. Is it pipeline? It really doesn't matter, right? And then that layer on top of it, right? We're democratizing AI across the stack, right? And that goes everything from the SQL embedded AI to fully agentic AI experiences that were helping you make accessible, right, to not just the builders in your org, right, but to all the users as well. And ultimately, that's what's going to help you, not only drive efficiency, right, on your data side and through your stack, but also have that business impact, right? We're making sure that everything, right, top to bottom is designed for that better economics faster development for you and overall total -- lower total cost of ownership. All right. So let's go into some of the amazing announcements that we made. And again, this is just a slice, right, of what we're doing. Next slide. Okay. So on that kind of base layer, right, how are we making sure when you bring your data in, right, that you feel confident, right, that we are giving that lower total cost of ownership. So here, we're making advancements that are really going to help you, as partners, not only simplify, right, your operations and governance, but free up those resources so that you can focus on more better value-added things, figuring out what the heck is going on with AI right now. So the very kind of basic level bullet point to. I'll start there, right? We've unveiled the standard warehouse generation 2, and that's going to give you 2.1x faster analytics performance. It's going to accelerate the insights that you can bring to customers, right? We all love faster, faster is better, it's awesome. But we all know right managing warehouse is not always the most fun thing to do, rightsizing that correctly, making sure you have that right-sized compute to scale. And that's why we're also introducing adaptive compute, right? This makes warehouses even easier for you to use. It's taking that burden of platform management that nobody really loves off of your hands, so that you can deliver faster results also at a lower cost. Really exciting. It's in private preview, right? You should go and you should check that out, right? Similarly, simplified ingest pricing. A lot of you are using Snowpipe to bring that data in. Thank you, do more of it, right? And we have more coming up about how you can even bring more data in and we're making lower pricing for you, right? It's going to give you faster data onboarding again, better price performance. And so you can be confident in that overall total cost of ownership. And then, you've got this data, right? We've got you the good price performance. How do you make sure, it's secure, it's governed, right? The right people are having access to the right things, right? And that's where our total horizon, right? What we're doing on the catalog, but also with the governments is going to be so helpful, right? We're doing enhanced interoperability, AI-powered security and governance in our Horizon catalog. That includes a copilot for Horizon catalog as well as AI-led monitoring, right? So we're taking, again, across the board, more of this off of your plate, right? So you don't have to focus and you don't have to worry as much about kind of the performance and the compute in the warehouse layer, right, and even the governance. We're helping bring AI. We're helping bring that lower cost right? So you can focus on, right, the next stuff? How do you bring in more interesting data, your unstructured data? How do you put it to use? All right. Next slide. All right. That's what we have, smarter infrastructure and governance. Next slide. Turning the slides back up. We've got more to talk about. All right. Next one. Okay, here we go. Okay. Accelerating development. All right. Tools for builders. This is -- I was one of the co-founders of Streamlit. This is where I get really excited about. How do we make it easier for you to bring that data in and do interesting things with it. So we're investing in so many new development tools. And that's really going to speed up your project delivery, help you build more differentiated offerings and drive innovation. Look back there. Don't give away the good stuff on AI. Okay. So first, right, you've got a lot of data, right, that's sitting in SaaS tools, it's sitting in lots of other places. You want to bring it together, right? You want to get that great price performance that I was talking about. But you're going to want to do interesting things with the 2 that we're about to show you with the AI side. So getting that connectivity, managing things in, open flow is really going to help you with that, right? Really, it facilitates this open interoperable architecture, right, moving your data of choice around your data lakes and lake houses, it makes it really, really easy for you to adapt to all these new industry standards, like iceberg, right, and bring that data in or write back, right? So unifying structured, unstructured, backed streaming data into a single platform, all with these kind of connectors out of the box, you can get all of this data in. And then we have workspaces and built on top of workspace we have DBT projects. This is 2 exciting announcements. Workspace is a modern development environment. We'll show you that in a minute, but it makes it really easy for you to edit things. Starting with SQL files and then we're adding in notebooks and streamlits that will all be together kind of 1 central command system for you to be working with your data in your code. And the DBT projects, you can run DBT now natively inside of Snowflake, right? So if you're already using DBT, this is great. You should go you should try it out running your pipelines there. If you're not, you should try it. DBT is an amazing open source tool. And speaking of open source, right, in Postgres, we announced a major acquisition at some at the Crunchy data acquisition, go and check it out. This is going to make it easy for you to do transactional workloads inside of Snowflake. So go take a look at that. We won't demo it today here, but you should check out the recordings on that. And then what I will give you demos up in just a second, Cortex AI SQL and semantic use, right? So these are 2 important building blocks that are going to help you build even more interesting things. So I'm not going to say too much about them now because we'll jump into a demo in just a second. All right. Let's round out some of the big announcements, and then we'll go over to the demos. Last slide. All right. So -- you've got your data in, you've got this solid foundation, right? You transform it, you're ready to put it to use, right? How do we help you put it to use? Well, first, we have an agentic framework, right, with Cortex agents, and we have Snowflake Intelligence, which is an interface that you can be giving to your business, right, to surface those agents, right, and put them to use. -- right? So this is really exciting for all of you that are really embracing right, these new AI capabilities who want to go to this next wave of intelligent applications. These are 2 really powerful things both to build the agents, right, and to surface them to your customers. Then back to the builders, right? We're giving you a lot more tools in terms of co-pilot. We'll show you in-line copilot and workspace, which helps you find data, edit it, give suggestions, fix, explain your queries, again, really awesome ways that we're helping make you more productive. And then core technology extensions are a really cool thing that we're introducing in the Snowflake marketplace that allows you to share your Cortex search services, right, it's private listings organizational listing. So that's going to make things like RAG, bringing in your articles, market research, right, into your agents more possible and easy, right? Again, what we want to do is, where we're bringing in all these models or bringing all these AI capabilities so what makes it really easy for you to build agents and apps that are going to deliver the insights that you have from all that unstructured data, right, that structured data that you're bringing together, right, for that next level of insights for your company. All right. Let's switch over and I'm going to talk even faster as we go through some demos. And again, we won't get through all of it. So you're going to have to go check out, right? Some of those videos online. All right. So let's go here super quickly. This is open flow. What does open flow do? Well, there's tons of different connectors, right? Just look at all of the connectors that we have here. If you don't see a connector that you want, this isn't even showing all of them, you can go and you can make your own with NiFi, right? It's a great open source project. And so when you have them, then it's really easy to set up these run times in these deployments. So we've got run times here. We're showing 1 for Kafka, Polestar SharePoint, rights go to bring in structured, unstructured, batch streaming again, right? It can do it all. And then you can have this running, you can have it inside your VPC, right? It could be on AWS or you could be running it yourself, right, in Snowflake, on Snowpark container services. We're giving you a lot of flexibility. I won't go into the full wave that you make these, but it's a lot of great things that you can do and it'll help you bring your data into Snowflake. All right. So you brought your data into Snowflake, right now, what can you do with it? Well, one thing is, right, you've got your DBT pipeline. That's going to help you, right? Makes sense of a lot of this transformation. So you can see here, I'm in a workspace, little one, we call Summit Fest here. You can see I have a DBT project. It's actually connected to [ Git ]. You can see some of the changes. I can push them if we wanted to take a look at them right now. So here are some of the dips right here in terms of the side-by-side, right? But if I go back, oops, oh, no, accidentally clicked something. This is how you know demos are live folks. Okay. But going back in here, we can see we've got my models, we've got [ YAML ] files, right? I can run this -- it's going to compile, right? We can see the DAG from the last time I compiled it, right, all here, right, inside of Snowflake, right, inside of a workspace, making it easier -- even easier than ever, right, for you to manage your data. All right. What else do we have? All right, AI SQL. So I skipped over this later -- earlier, but this is really, really cool. So I'm running this in a notebook here. And what I want to show you, right, is how you can use Cortex AI SQL to embed generative AI directly into your queries, right? That's going to help you analyze all types of data with just the familiar SQL syntax, right? So here, we're going to show -- we're trying to find some customer issues across text, across image, across audio data. That would have been really, really impossible, very hard in SQL before. But here, right, using AI complete, right? We are able to do this with some multimodal prompts, very little amount of code that we're doing, that's going to allow us to consolidate that data across all of these different formats and start to put it to use. And then as we have that, we're going to also be able to not only consolidate it right across the text, the image, the auto, but we're going to use the power of AI to semantically join those customer complaints to the solutions, right? Again, we're just using AI here in SQL. It's allowing us to work with this data. It's allowing us to new amazing things so that we can get these aggregated insights across all of these different types of data, right? And we're doing all of this in just a few lines of SQL on the Snowflake platform, getting direct access to the best frontier models that we have, right? This is an amazing kind of ability that we're providing you out of the box. It's going to give you amazing productivity gains. And if you don't want to take my word for it, check out this amazing benefits that we're measuring here for you, in terms of performance and cost benefits, right? Sometimes 3 to 7x performance benefits. It's really amazing, and you should go and you should try that out now. All right. What else do we have? Well, let's say, you're saying, well, you know what, that's great. But I actually want to be doing my own ML. Well, notebooks now, we have them, they're GA, you can use them on containers. You can be using them with GPUs. Here I am, I'm running something in the background here. We're predicting diamond prices. I've been running this at GPU for a while. You can be managing your CPU, right? You can be looking at your memory, right? You can do a lot of really sophisticated things now, right, both with our ML platform, right, and with our notebooks, right? You can do a lot of these things too with Copilot, which if we have time in a minute, I'll show you as well. But supercool ways we can share these now, getting so much you can do. All right, semantic models. I promised we would talk about that, too. All right. So hopefully, you know what a semantic model is, and you know why it's important, but semantic layers help you unlock these AI-powered analyst experiences, right? You can create consistency across your AI and BI. It's really important as you start to build out these agents in these models. And we can also take the semantic models out of the multiple BI tools that you use right now and that every customer has and you move it into a single day or later, right? So we can see here, we've got our tables and schema. And now, we can just very quickly create a semantic view. So here's our DDL for the semantic view. We're just doing this in notebook, we're doing it in SQL. Very easy stuff that you're familiar with. And then here, we can describe it, right? So since this is stored as a native object, we can describe it just like we do a table or a view, right? Is that easy? So now we can go over to Cortex Analyst, and we can ask you a question, like, I'll copy this question here. What are the top 10 brands for the books category in the state of Texas. So if I go over here and I ask that, right? Now we're going to use that semantic view. We're going to use Cortex Analyst, it's going to help us interpret that question, generate the SQL, run the SQL for us, right, and get an answer. So this is really cool. I'm starting to talk faster because of all of the things I still want to show you, right? So here's the semantic query. We can look at the physical query. We can go back here. If we ran it here, right, in SQL, you'd get the exact same result, right? So this is a new and very powerful thing that all of you are going to want to explore. It's going to give you a lot of power, especially as, right, you move into more of these AI things. Okay. A couple more things, right? This is a workspace again. Again, amazing things that you can do, we can pull charts in for you, you can filter them in, you can do these fast. You can look at things side by side. Let's do a new query here. It can find things for you. Look at this, I can say, fine, my COVID-19 data set and give me a sample, right? It's going to go. It's going to use our search under the hood, look across the data sets that I have available. It's going to pull that for me, right? Give me the answer. I can go ahead. I can run that, right? It all works. Again, we're just weaving all of this AI things in for you. So whether you're a builder, whether you're somebody building those semantic views, whether you're trying to govern your data, it's all there for you. And last but not least, right Snowflake Intelligence. This is how we help you bring these agents together. Those semantic layers, everything is coming together here, right? So we can ask a question here. We can ask directly in chat. These are things that you can be giving to the business users in your company, and it's going to do everything that we just showed you. It's building all of that together, right, the agents, the semantic layer, the search, all of that is coming together in order for it to reason about it, to go into your data to actually provide you these direct answers for. So I don't know if I'm out of time yet. We'll keep it on this while I kind of close as it's doing the reasoning and thinking about it. But really, I hope you've gotten from this that Snowflake is no longer, right, just where your data lives, right? It's where your AI applications are born. With these innovations, you now have that toolkit at your fingertips to deliver the smarter, more personalized and more scalable customer solutions, right? And the opportunity is really just wide open for you as partners, right? Those who move the fastest with these new innovations are going to have so many more things that they can bring to their business, to deliver more value to customers and to deepen those strategic relationships. So please go online, check out the Summit replays. There's even more better exciting things, go to our solution center, see how some of these things are coming, turn on those previews in your accounts, we're really excited to see what you build.

Thanks, Amanda. Really very, very cool demos. Yes. We have a lot of questions that comes at the Q&A. So can you just keep coming -- keep asking, we have SA that's helping with the solutions, and the questions. So go ahead, and thank you so much, Amanda, for all your sharing today.

Thank you. It was exciting. I look forward to coming back to Asia soon. I always love everything that's being generated here. But I will say good night since I'm in the U.S.A.

Yes. Thank you so much, Amanda. Bye. Hi team, we love to have your feedback. So give us your poll in about 2 more minutes, we will close. So we also have up and coming our Snowflake World Tour. So happy to get your polls. And if you are interested to join our sponsorship of our world tour that's coming to our region, do check it out and share your interest as well. And next up our poll, which is what you guys are feeling. Really, really appreciate all the feedback that you have given us to make our show even better the next one. So we'll come back definitely in quarter 3 and quarter 4. So on a quarterly basis, stay tuned for all the updates that we will share with you guys. So I'll be online for another 2 more minutes for every one of you to fill in your polls, and we really, really appreciate any feedback from you. Hope, all of you had an amazing time. And all the partners, I really appreciate you taking your time to spend this morning with us or afternoon. Yes, I really appreciate it. We will close webinar in about 1 minute's time. Team, thank you so much, and we will end house. Thank you.