Snowflake Summit 2025 | Mona Chadha, AWS

Clips
News
More from Snowflake Summit 2025

Mona Chadha

Director of Category Management

AWS

From prototype to production: How AWS and Snowflake are scaling gen AI

Data integrations are emerging as the critical link between generative artificial intelligence experimentation and real-world deployment, according to Amazon Web Services Inc. and Snowflake Inc.The two companies have joined forces to help customers move from proof-of-concept to production. Their next focus: unlocking the power of agentic AI and automation.Mona Chadha of AWS discusses Amazon’s partnership with Snowflake.“We’ve seen that with generative AI, everyone’s using some component of it, whether it’s with a large language model … or creating these chatbots across different use cases,” said Mona Chadha (pictured), director of category management at AWS

play_circle_outline GenAI trends in 2025: from experimentation to production, deliberate use, customer challenges, augmented teams.

play_circle_outline AWS and Snowflake partnership: data management, integrations, customer enablement, industry focus, ROI.

play_circle_outline Maximizing AI Success: Thoughtful Data Strategy, Team, Sources, Security, and Compliance

Info
Transcript

Mona Chadha, AWS

Mona Chadha

Director of Category Management AWS

In this interview from # Snowflake Summit 2025, Mona Chadha, director of category management at AWS, joins theCUBE’s co-hosts Dave Vellante and John Furrier to unpack how Snowflake and AWS are accelerating enterprise generative AI adoption on the AI Data Cloud. Chadha explains why an integrated Snowflake-on-AWS stack is becoming the connective tissue for building, scaling and governing intelligent systems – eliminating data movement while preserving strict security and cost controls.

The conversation dives into the nuts and bolts of putting gen AI in... Read more

explore Keep Exploring

What has been the evolution in customer attitudes and approaches towards generative AI over the past couple of years? add

What integrations does AWS offer with Snowflake to help augment customers' foundational data challenges? add

What are some top advice for implementing a successful data strategy? add

bolt Powered by CUBE AI

Mona Chadha, AWS

search

Rebecca Knight

>> Hello everyone and welcome back to theCUBE's Live coverage of the Snowflake Summit 2025 here in San Francisco. I'm your host, Rebecca Knight, sitting alongside my co-host and analyst, Dave Vellante. I would like to welcome Mona Chadha, director of category Management at AWS to the show.

Mona Chadha

>> Thank you. Thank you.

Rebecca Knight

>> Welcome, Mona.

Mona Chadha

>> Thank you.

Rebecca Knight

>> Thank you for coming on the show.

Mona Chadha

>> Oh, thank you so much. I have to say I love this event because it's where all the data and analytics providers are, and it's like I can get my job done all in one shot here. It's so great.

Rebecca Knight

>> These are your people, these are your people.

Mona Chadha

>> These are my people. Yeah.

Rebecca Knight

>> So let's talk GenAI.

Mona Chadha

>> Yes.

Rebecca Knight

>> 2025 really seems to be the breakout year for GenAI for the enterprise. From where you sit, why don't you just give our viewers a broad brush strokes of what you're seeing and what trends are emerging?

Mona Chadha

>> Yeah, it's funny because GenAI, I would say even two years ago, it was all about experimentation and a lot of what our customers were doing was trying it out because they didn't really know how to integrate it into their enterprise applications or into their enterprise workflows. And I think what we've seen is that with generative AI, everyone's using some component of it, whether it's with a large language model like an Anthropic or creating these chatbots across different use cases, customer service, et cetera, sales. Really helping to accelerate the customer experience. What we found is that now they're moving away, a lot of those customers that were trying these experiments are now implementing more production, and that's been actually pretty interesting because it's gone out of the whole, "I don't know what to do with GenAI phase" to "This is exactly what I have now a better idea, and I can get prescriptive guidance for a variety of analytics providers, a variety of cloud providers on how I want to help solve my challenges." And I think that's the other key thing that has evolved is everyone's thinking around generative AI or just around AI in general. It's more deliberate versus before it was, "I don't really know which, I'm just going to experiment. I'm just going to try different use cases," but not really clear on how it's going to help them solve a customer challenge. And now it's very deliberate. "Here's the challenge that I need to solve. Here's how I need to integrate it into my standard workflow. Here's how I needed to help accelerate my business." And then also I think another key component is people are not just looking at it to replace their employees or anything like that. And I know there was a lot of concern about that, but now-

Rebecca Knight

>> There still is.

Mona Chadha

>> And there still is, but I think now what's happening and what we're seeing is your resource and your human resource is very precious, and the skills that they have are significant and sizable, and they're like an algorithm. They're the best algorithm out there, especially when they built up years and years of knowledge. So being now able to take some of that, I would say, I don't want to say undifferentiated, but these are really still strategic capabilities and building it into generative AI and agentic coupling that with agentic to get your automation going only helps augment your teams. And I would say the productivity and the efficiency of your organization. And we're seeing a lot more of that happening across different industries like financial services, healthcare, manufacturing. That's been huge in manufacturing to really help that. So we're starting to see more of the POC to production, the deliberate use of AI overall between agentic as well as generative AI. And then really customers asking for, "What are those best practices? Give us some ways in which we can do AI not just responsibly," that's definitely a key component, "but how do we do it right?" And that's some of the advice that we're giving as well as a lot of the people here are giving at Snowflake as well. We're a great partner with Snowflake in augmenting customer's AI strategy.

Dave Vellante

>> What does that mean? Getting it right? It means getting it profitable?

Mona Chadha

>> Well, definitely ROI is a big component of it, and I don't know that everyone's got that right. I think that's moving in that direction, but it is. The way that we sort of look at AI in general is thinking about it more holistically. And then a couple of things is thinking about stuff like, well, what is the use case you want to use? Don't just use AI overall because a board asks, "Hey, how are you using AI, generative AI, agentic or whatever," but really be deliberate on what you're using. Then be thoughtful about security, be thoughtful about compliance, and then be thoughtful about your data strategy. And those are elements that start to get it right. Leverage your teams, make sure that you're really clear on what that ROI or what that business outcome you're trying to solve. And that's really what it's about, about getting it right, is adding all of those elements to the equation.

Dave Vellante

>> Amazon Web Services specifically, the impact that it's had on our industry has been quite remarkable when you think about it. People talk about always now talk about working backwards. You never heard that. Some people would talk about, we focus on a customer's customer obsession. You never heard that phrase, two piece of teams, remember that?

Mona Chadha

>> Yeah. Day one.

Dave Vellante

>> There's no compression algorithm.

Mona Chadha

>> We still have that. We still use that.

Dave Vellante

>> And the other thing would say a lot was "We're comfortable being misunderstood."

Mona Chadha

>> Yeah.

Dave Vellante

>> One of the things that's often been misunderstood is your relationship with Snowflake. It's like the poster child for co-opetition, right? Because the Redshift, my guys might love Snowflake, but Amazon sure loves Snowflake because you got great partnership. As you said. Another kind of misconception is, oh, "Amazon's closed. It doesn't like open source." Now you look at what you're doing with Iceberg.

Mona Chadha

>> CP, yeah, Iceberg.

Dave Vellante

>> Right, then OCP for sure. And then another one is, oh, just primitives. It's all primitives. We saw it reinvent this year with SageMaker integration.

Mona Chadha

>> And Nova, right?

Dave Vellante

>> And so of course Nova as well. So I guess one of the leadership principles is be right a lot. So that's hard to do, but I wonder if you would talk about long-winded question. Wonder if you would talk about your Snowflake relationship?

Mona Chadha

>> Yes.

Dave Vellante

>> Where you see that going, how it fits into your vision of where Amazon and Snowflake are headed?

Mona Chadha

>> Yeah, and here's the thing. Snowflake and AWS have a over a decade partnership. And yes, there were components of it where there were some overlapping capabilities, but it's so much richer than that. And I think the key thing is that we came together really trying to solve customers foundational data challenges. And Snowflake really is that data platform where we found a lot of our customers are using Snowflake. And as a part of it, what we looked at was the integrations and having those integrations with a variety of AWS services that help augment your Snowflake experience and specifically data. What we are trying to do and what we have been doing for the past decade working with Snowflake is really collaborating on augmenting capability, augmenting generative AI. We have over 50 different integrations across a variety of different AWS services, from SageMaker to Bedrock to Amazon Q, and being able to really help Snowflake also differentiate their data foundational strategy. And the key component, and I'll give you a core example there is where you want to give your customers the variety and selection for your LLM models or any foundational model. And so as Snowflake looks at how customers want to leverage models, they use Bedrock as that sort of pipeline into all of that flexibility and all of that variety, all that selection. And so as an example, Anthropic, the way to use Anthropic with Snowflake is through Bedrock. They do the same thing with Nova. So we've seen just this, the integration that we have with Snowflake is really kind of unparalleled with other ISVs, and it shows the proof is in the pudding with our end customers like BMW, Allianz, where they're really looking at building these capabilities that are first of its kind across a variety of different industries. So capital markets, the work that we're doing with Epic, so any customer that uses Epic, the integration that we have here with variety of different SI partners with Snowflake and AWS.

Dave Vellante

>> Talking about the healthcare aspect.

Mona Chadha

>> Yeah, the healthcare, yeah, Epic. So I think the power of us together, it's so impactful for our customers that when we saw that with all the integrations that we had, and then the partnership that we have now spans across SIs and ISVs. So Snowflake brings this network effect of these ISVs that are attaching their BI tools to it so that you're now able to parse through all that data, make sense of it, and then ultimately accelerate your business outcomes.

Rebecca Knight

>> One of the things you were saying is that this is the year that organizations are getting much more intentional about how they're using AI.

Mona Chadha

>> Yeah.

Rebecca Knight

>> And one, it's because they want the ROI yesterday, but also they want to actually make sure they're solving problems here. But getting from pilot to production can be really difficult. What are some of the pain points and roadblocks that they're coming across, and how is AWS and Snowflake working together to solve them?

Mona Chadha

>> Specifically around data. I think that's the key thing, right? Garbage in is garbage out, and you have to make sense of all that data. And I think that's specifically where AWS and Snowflake are helping customers manage that and manage their data foundational strategy. I mentioned the work that we're doing together with the integration of Bedrock and then also integrating Amazon Q. and also the work that we're doing around helping customers take their data and be able to integrate it into Iceberg and into data lakes and making sense of that across structured and unstructured data. So a lot of our customers need to have, they need that initial guidance of how you're going to build out your data strategy. And that's I think first and foremost of how we're building that out. Then you can overlap the capabilities of Snowflake to really help deliver a holistic strategy around agentic and around, I would say, around agentic, and then figuring out how do you automate those specific workflows in your environment so that you can run faster, basically? So I think that's a key thing. I think the other thing that we've been doing is really working together to make sure that we have delivered... You have a plan to deliver a POC to production. And so that's a key thing where people take that for granted and they don't actually come up with a thoughtful plan. So we've invested time together in creating workshops and creating customer enablement across different industries so that we now have kind of templates that if you're in financial services, you're in capital markets, you're trying to migrate Epic data into your Snowflake environment, how do you go do that? Now we kind of have these templates across the different industries and working with SIs to deliver that so that it is very intentional. You have a map of where you want to go and how you're going to get there. And a component of that is also the ROI. What sort of ROI should you expect? So creating these, not just technical sort of architectural frameworks, but also the business framework to say, here's how you're going to drive your ROI and what you should expect. And so all of that between the data foundational strategy, the industry focus, and really helping navigate that to then the customer enablement, those are sort of the three things that we've been doing to really help accelerate our use cases.

Dave Vellante

>> So last year was all about kind of experimentation.

Mona Chadha

>> Yeah.

Dave Vellante

>> This year, it's production. You had said maybe not everybody has the monetization right.

Mona Chadha

>> ROI. Yeah.

Dave Vellante

>> Is that what we're going to be talking about next year, is that we are scaling that production and we've figured out the ROI piece?

Mona Chadha

>> Well, I think the way that you look at it's that never goes away. We're constantly going to be in that mode because if you look at the evolution of AI, last year it was about generative AI. This year it's about agentic, but they all layer on top of each other. And so as a result, you're going to get ROI from different kind of capabilities. And ultimately now we're in a mode of how are you going to get that autonomous ROI? How are you going to get that from automating some of your tasks? But then how does that relate to your employees? How are you going to upskill your employees and what does that look like? So I think the ROI conversation may never go away as long as we're continuing to innovate. And I think it's okay as long as you have a good direction on how you're trying to, which customer challenges you're trying to solve and have that priority and that focus.

Dave Vellante

>> I like that answer. I used to get ROI from putting up a website now, not anymore. It's like table stakes.

Mona Chadha

>> Yeah, .

Dave Vellante

>> It became e-commerce.

Mona Chadha

>> It's going to evolve. Your ROI evolves. Yeah.

Rebecca Knight

>> So I want to ask you about what success looks like and when a customer has succeeded with AWS and Snowflake, what is true behind the scenes? Is it the technology? Is it the organizational structure? Is it the mindset and the approach and the culture that, what is true?

Mona Chadha

>> I really like this question because that everyone would say, especially probably you ask everyone here, it's the technology, it's the integration, and it actually isn't that. It really is the last thing that you said, which is changing the mindset, changing the culture of your company. When you can get companies, you can get people, not even companies, when you can get individuals to really embrace change, that's success and do it quickly like this. And it's not like, let me think about it. I look back at my career and think about how we went from outsourcing to offshoring to then the cloud and oh my God, the mindset. Even today, people are not 100 percent on the cloud. And so these are now it's evolution of thinking, but I think success looks like when we have changed the mindset of people to say, "You know what? Here's what I'm trying to drive and the technology is an enabler to this, and I'm not going to be blind to saying I need to keep my data all on-prem, or I need to keep it here. No, it's how I'm going to be able to move quicker and faster for my end customers." And that to me is the change of the mindset, the cultural thing, and not having that fear of your job going away or that fear of, "Oh my God, how am I going to learn something new?" That to me, that's success when people are like, Nope, that's not the thing, because I can use this technology and I'm not afraid of it. I can use it and I am asking the right questions. And I think once we get people to start thinking more about, it's the question that you need to ask, that's the right thing versus the technology that you implement.

Dave Vellante

>> Well, and then of course there's always public policy and regulations and things like that because one could infer from your comments that a path to that acceleration is the cloud and the fullness of time. You believe that most workloads are going to be in the cloud, but then you have this sovereign AI and countries say no. And so that sort of forces companies like Amazon to say, "Okay, we're going to put a region here." So it's like GovCloud all over the world, and that's playing out.

Mona Chadha

>> Yeah, it is. And that is part of what we are doing with Snowflake as well. Because you're right, I think people's, the mentality won't change. And I think part of that mentality though is also because potentially bad actors coming into play and maybe there's been some experience around there. And that's why I think when you do build the technology, and this is where Snowflake and AWS have an aligned thinking on in many ways, and definitely in this way is that security, having security. So anything that we build has to have that security mindset and being able to protect that data. But I think ultimately that is the thing where we're looking at growing and building regions with, so AWS is building new regions, and we're like the sovereign cloud going into new regions like Saudi. And also just ensuring that we have availability of those capabilities for our end customers there. And Snowflake is working with us on that.

Dave Vellante

>> You guys have been in the region, in the Middle East for a while, I know, because we did two CUBE gigs in Bahrain.

Mona Chadha

>> Yeah, Dubai. Yeah, Bahrain, that's right.

Dave Vellante

>> With Theresa Carlson back in the OG of GovCloud.

Mona Chadha

>> Oh yeah, that's right. Exactly.

Rebecca Knight

>> Exactly. So I want to end this conversation where we began. 2025, the year of break out GenAI for the enterprise. What is your top advice for the companies that, as you say, want to get AI right?

Mona Chadha

>> Yeah, I would say my top advice would be really be thoughtful about what use case, what are you trying to solve? Number one, I think two, be very thoughtful about what is your data strategy. Do not take that lightly. Ensure that you have that, ensure that you have the security, the compliance all embedded in there. The other thing is that make sure that you're bringing your team along. I think that might even be the first thing that you do is you have a team that does have that mindset, that growth mindset of this is how I want to evolve, and ensuring that they're implementing the right capabilities, whether it's agentic, it's generative AI, whichever models, because there's a plethora of models. And so you really have to now be thoughtful across which models you want to use in making sure that the data going into those models, there's veracity of that data and what's coming out is very clear. And so testing those out. So definitely I don't think experimentation ever goes away. You need to do that. But I would say you need to kind of be very more thoughtful about what your data strategy is, integrate it. Don't just use data from one source. You need to also supplement that data from multiple resources to again, ensure that veracity of that data. And then you got to be responsible. You got to make sure that you're doing the right thing for your end customer and yourself and your team of people. So I would say do that and then check in. And what I mean by check in is you're going to test, you're going to experiment. You're going to see what that looks like, measure the ROI and keep iterating. And I think that's the biggest thing is keep iterating on what you're doing to improve. And you have partners like AWS and Snowflake and the hundreds of partners that are here that can help you do that.

Rebecca Knight

>> Excellent. Great advice. Words of wisdom.

Mona Chadha

>> Great, thank you.

Rebecca Knight

>> Thank you so much, Mona, for coming on theCUBE.

Mona Chadha

>> Thank you so much.

Dave Vellante

>> Good to have you, thank you.

Rebecca Knight

>> I'm Rebecca Knight for Dave Vellante. Stay tuned for more of theCUBE's live coverage of the Snowflake Summit. You're watching theCUBE, the leader in enterprise tech news and analysis.