Databricks Data + AI Summit 2024 | Rik Tamm-Daniels

Clips
News
More from Databricks Data + AI Summit 2024

Rik Tamm-Daniels

GVP, Technology Alliances

Informatica

Navigating gen AI challenges: Informatica and Databricks partner for data governance innovation

Although generative artificial intelligence may offer a variety of productive use cases, it’s not without its downsides. One annoying downside is generative AI’s ability to hallucinate results, spitting out inaccuracies and useless code. AI governance and well-defined blueprints help stop these issues.Informatica Inc.’s partnership with Databricks Inc. is bringing innovative products, such as DBRX and IDMC, to the ecosystem, helping with both data management and governance and allowing for seamless and accurate AI integration, according to Rik Tamm-Daniels (pictured), global vice president of ecosystem alliances and technology at Informatica

play_circle_outline The relationship between Informatica and Databricks

play_circle_outline Informatica's focus on GenAI and bringing experiences to the edge

Info

Rik Tamm-Daniels | Informatica

Rik Tamm-Daniels

GVP, Technology Alliances Informatica

>> Good afternoon, CUBE community and welcome back to Databricks Data and AI Summit. We're here in our hometown of San Francisco, which I am fabulously delighted about. My name's Savannah Peterson, joined by my co-host and CUBE co-founder, John Furrier. John, what's up, man? Afternoon, we're grooving.

>> It's grooving. It's rocking. The show's packed. The overflow from the keynote almost filled up the entire cafeteria room, which is massive. So a ton of people here interested in what Databricks is doing again. AI is driving everything.

>> A ton of people here, a ton of people across business units too, which is a really exciting thing. We've got developers, we've got leadership, we've got executives. We've also got a fabulous guest joining us right now who's been on theCUBE, maybe even more times than I have. Rick, thank you for coming on the show today.

>> Thank you. Thank you for having me.

>> What keeps bringing you back to punish yourself with John and I?

>> I love talking with you guys.

>> Rumor has it, you were running the show at Informatica with three cameos not that long ago, which is awesome. I want to start off by noticing the first thing I noticed about you when you walked up to the desk today.

>> Yes.

>> You are wearing a shirt that implies you may have received some accolades here. Can you explain your shirt?

>> Yes. We decided to be very subtle about this, but we are really excited this year that Databricks has recognized Informatica as their 2024 Data Integration Partner of The Year. So it's just huge validation of the growth we've seen together. Last year, Databricks named us one of their top three fastest-growing technologies. And now to receive this, just reflect with that momentum, a ton of innovation as well. That's a lot of what we're talking about here this week.

>> What's the competition like? How many people did you have to beat up for that slot, do you think?

>> Oh, there's a lot of players out here. There's a lot of folks in integration these days and-

>> Have you walked by their booths with your chest out a little bit just to...

>> No comment. No comment at all.

>> I love it. Well, obviously a very deserved accolade. So what are the types of solutions that you're providing partners that everyone's so in love with?

>> Well, so one of the biggest ones that we rolled out for the Databricks Data and AI Summit is we call our Generative AI Blueprint. And we've been rolling these out across multiple ecosystems, but the idea is how do you take all the capabilities of our intelligent data management cloud that provide trusted data foundation and bring those into the world of generative AI to enable what we're calling enterprise-grade GenAI apps?

>> Enterprise-grade. We hear that a lot. What does that mean exactly? What sorts of check boxes does that have to tick to qualify?

>> Well, there's a few key requirements we look at. One is, everyone knows that one of the big potential downsides of GenAI is that models hallucinate, right? They make things up and in the enterprise, that does not work, right? You need consistency, you need accuracy. And so grounding becomes a critical capability, and we found that things like our master data for Customer 360, supplier product really provide a good foundation for that. One of the areas is contextualization, right? A public model doesn't know... When you talk about your customer, is that a company or is it an individual? That's very specific to your domain. So bringing the governance, the intelligence about what your semantics are of your business, bring that into your prompts, to enrich your prompts, enrich your summarization, creates a lot. And then we also look at things like security, making sure that users only see the data they're allowed to see. That's critical as well. And ultimately creating the infrastructure, the scaffolding. How do you do that in a very scalable way, a no-code way? And the great thing is our platform is really well-positioned to deliver on all those elements.

>> So one of the things this show's highlighting is, again, the conversation we've been having on theCUBE and we had at Informatica world was AI really has to nail the governance upfront. That has really been a big thing, all the traction has been. And governance used to be talked about, but certainly you guys, that's your business.

>> Yep.

>> But it's now mainstream around that because of the horizontal scale of data and now that developers are involved. So it's not just set the data up for data engineering. It's like it's impacting the app web cycle and development cycle. Why is that happening and what are some of the things you're seeing around where people are in that progress of their journey in rethinking or just deploying more governance?

>> Yeah. Well, I think a lot of the same underlying factors that drove the need for governance in the analytics world have just been amplified in the GenAI world. And I think in a lot of ways, GenAI experiences have potential to be magical experiences. Right? And the problem with magic is you're wondering, "Well, what's going on behind the scenes here? How do I know-"

>> Are you or are you just loving it because of the context?

>> Exactly. But enterprise recognize that if they lose trust in these applications very quickly, it's not just visible to the IT folks anymore. It's visible to all these business users. It impacts them. So the downside of not having governance is much, much bigger. It's much bigger risk for companies to not have that grounding, that accuracy, that transparency in their generative AI applications.

>> If you don't have governance, it's not going to work.

>> Yes.

>> It's a non-starter at this point.

>> Absolutely.

>> Especially with the scale and everything else. Let's talk about no-code and low-code in the data pipeline support of Databricks and end user data integration and transformation?

>> Absolutely.

>> What can you tell me about it? How do you help that?

>> Well, yeah. So Informatica has very much been focused always on the no-code user, configuration-based, visual drag-and-drop experiences. One of our big announcements this week is we call it our native SQL ELT. A long acronym there, but the basic idea of using visual definition of your data pipelines that we push down and run natively inside of Databricks. So we use the compute power of Databricks, the extensibility with their... They have a lot of specialized functions these days that really lets customers take advantage of the full platform. But again, take the scalability, maintainability of that no-code experience.

>> It sounds like a pretty lovely turnkey solution. Everything I'm hearing in what you're saying is that easy button that everybody's looking for. You're making it all sound like this is going to be smooth and dandy. Are there any concerns or roadblocks as you look out in the future of risks for customers or things people should be aware of that you could advise them to avoid?

>> Yeah, absolutely. I think one of the big things is just like any initiative with data and analytics or AI, you really have to understand where are you headed, what are you trying to achieve, what are the outcomes you're going to have? So as much as this is a technology problem, it always comes back, there is this people process side of it as well, and setting expectations correctly with your consumers is critical. If the expectations are way out of whack-

>> That's a really good point. So let's talk about that just for a second, because I feel like customer expectations, and correct me if I'm wrong, they've got to be all over the place right now. So how do you work with your customers to get them to realize what's actually going to be a tangible outcome today versus what's going to be a tangible outcome say in the future?

>> Yeah, I think a lot of it is... And we have a great partner community and ecosystem as well. A lot of the system integrators and GSIs, they are critical to that, because they can come in and help the organization with how are you going to adopt to this, what's the change management around it? And that's been a lot of the conversation is, "Hey, before we go and invest in this big initiative, let's understand, are we going to get a 50% improvement, a 10% improvement? And then how are we going to measure and track that on an ongoing basis? How do we build the practice around it." Just like we've done around analytics and reporting, those quality controls, how do you do that in the AI world?

>> Talk about the relationship with Databricks. Obviously you guys are on the ecosystem, Partner of The Year for data integration.

>> Yep.

>> BFFs.

>> You're starting to see the clients, and Ollie said it on stage today, the data estate's been fragmented. Does that problem go away? There's always going to be multiple vendors in there. You guys coexist a lot with Databricks and other platforms. What is the customer perspective on this? Can you share one, your relationship with Databricks, and how are the customer's viewing it?

>> Well, I think Informatica occupies a very unique position. I think, as you guys know, we're the only platform I'm aware of that's on four clouds out there. And we're recognizing there's a lot of innovation. Customers who want to try out different things, but one continuous thing is they want consistency in their data management, in their governance, no matter where they are.

>> Consistency is critical.

>> And that's huge for us. So a lot of customers are coming and saying, "Look, we want to move away from point solutions. We actually want something holistic. We want something that's scalable long-term," and that's worked really well for us.

>> That cohesion, that single dashboard, that single toolkit, that UI, that's really thoughtful. How often are you communicating with customers, with developers, with your community to keep that feedback loop relevant? What's that like for you?

>> It's continual, right? It's a really important part of... And especially we do on the ecosystem side, my part of the business, is having deep product alignment with Databricks on an ongoing basis. It's a big part of our approach because when we have new innovations coming to market, like we just announced the full verification of support for Unity Catalog across our entire platform-

>> Good timing.

>> Yeah, exactly, exactly. But that's because we're coordinated. We're aligned with them. We understand what's coming next and what's going to be most important for customers. And then here, customer conversations are absolutely critical. I love being here and getting to talk to customers, share the new announcements, get their feedback. It's a great environment for that.

>> Are there any customer use cases that get you excited personally? As much as I'm sure they get you excited as a business, but just as a human?

>> I love cases where you are transforming how people are working day to day. And actually, I think one of the potentials of GenAI is that for the first time, we really can start to see not just data being democratized, but data management being democratized. And that's a lot of what we're doing with our CLAIRE GPT and CLAIRE Copilot. I'm really excited for those applications. If you think about a supply chain analyst who's trying to do new sourcing and they're able to look at public data, figure out, "Okay, well, what are the right products that are exciting, getting traction, but which suppliers do I actually work with already and have relationships because they have their MDM data together?"

>> I think that's .

>> What's next in this evolution for Databricks, you think? You're starting to see the openness coming. What do you see them doing next?

>> Yeah, openness is a huge theme, right? When we look at the open sourcing announcement of Unity Catalog, and you look at what's going on with Tabular and Iceberg and the Iceberg adoption. So I think we're going to actually see some convergence of standardization, which is good, which is great for everybody in terms of skillset development, scaling what everyone's trying to achieve. So that is definitely one of the big themes, and I think it's going to be good for everybody.

>> I love the serverless.

>> Yes.

>> Yeah.

>> It reminds me of taking some friction out of the managing all these clusters. I fetched some data on it, Savannah, and the openness too-

>> Yeah, it was a big announcement.

>> It was a big deal. Yeah, it was.

>> I think that was probably the hottest takeaway from the morning when I was watching the Twittersphere in my car on the way home, or in the car that was driving me here. For the record, I wasn't watching this whilst operating the motor vehicle. Hopefully soon I won't have to operate a motor vehicle again. But yeah, I wasn't expecting that. I actually wasn't expecting a few of the announcements that came out today that were pretty exciting. Did you know everything that was coming?

>> We have a little view into that and-

>> Oh, that must be nice. All right, well, on that note, what's coming next then? What can we expect?

>> Tomorrow.

>> What can you tell us next? What can you tell us next? What's the future for Informatica, first? We'll go that direction so you don't have to be too naughty yet.

>> Okay, fair enough. So the big thing for us is, so we went GA with our Claire GBT and Claire Copilot. So the leveraging of GenAI inside our platform is a huge area of focus for us and customers came out of Informatica world so excited about that. We're demoing it here to the booth, showing folks what we can do with it. So I think that's huge. Where I'm focused as well is, again, bringing those enterprise-grade GenAI apps as well and bringing those experiences, and also starting to bring those experiences to the edge, so to speak, meaning at the point of need, the point of consumption. You look at, I think some of other ecosystem players, where they're embedding copilots and things like that. There's some really unique opportunities that are going to be presented as customers start to get to the true enterprise adoption.

>> Okay. I got a question for you because I was going back and forth with Sean and Dave a little bit yesterday about what I think is going to be the edge device that onboards the most people to adopting GenAI on a more daily basis. So I'm curious, what is your prediction for what's going to be that edge device?

>> So I'm not going to name a specific technology necessarily.

>> You can name what's better.

>> But I would say collaborative tools for me are definitely one of the hotspots, right? Because you have communities of interest that have common needs that are sitting together. They're collaborating already in these spaces, but now with GenAI apps and copilots, they can actually translate that to action and actually build the application experience. So that's a big one. I do think things like office, sitting on the desktop is going to be a big one. I use it already day to day. So I think those are the two big categories for me. And then probably further down the line would be embedded in specific applications, and there's a line of business applications and having the experience there as well.

>> All right, that was nice and vague, but also, you gave us a little bit of a picture there. I see what you're doing. All right. All right. I'm going to ask you one of my favorite questions to ask because obviously we have you on the show constantly. So I'll actually give it a timeframe.

>> Okay. All right. All right.

>> I'll let it extend out. What do you hope to be able to say when we have you on the show at Databricks Summit next year that you can't say yet today?

>> I would love to be able to spend the whole time talking about customer apps that are delivering enterprise value with GenAI. We're in the early days, but I just want to come on here and talk about, "We did this, we did this. Here's amazing examples where customers are just... GenAI is really... The promise has come to fruition." Right?

>> I hear that so much. So you're going to bring a customer up here-

>> Sure. Absolutely. Absolutely.

>> Next time and we can tell the story together?

>> Yep.

>> I think that would be absolutely fantastic. All right, well, this has been thrilling. John, thank you for your insights, as always.

>> Appreciate it.

>> And thank you for coming to hang out with us, Rick.

>> Always a pleasure.

>> Clearly, you are a CUBE celebrity now for over a decade it sounds like, and three times at Informatica. So I look forward to sharing the stage with you next time. And thank all of you for tuning in wherever you might be enjoying this thrilling coverage of Databricks Data and AI Summit here in San Francisco, California. My name's Savannah Peterson. You're watching theCUBE, the leading source for enterprise tech news.

Rik Tamm-Daniels | Informatica

>> Thank you. Thank you for having me.

>> What keeps bringing you back to punish yourself with John and I?

>> I love talking with you guys.

>> Yes.

>> You are wearing a shirt that implies you may have received some accolades here. Can you explain your shirt?

>> What's the competition like? How many people did you have to beat up for that slot, do you think?

>> Oh, there's a lot of players out here. There's a lot of folks in integration these days and-

>> Have you walked by their booths with your chest out a little bit just to...

>> No comment. No comment at all.

>> I love it. Well, obviously a very deserved accolade. So what are the types of solutions that you're providing partners that everyone's so in love with?

>> Enterprise-grade. We hear that a lot. What does that mean exactly? What sorts of check boxes does that have to tick to qualify?

>> Yep.

>> Are you or are you just loving it because of the context?

>> If you don't have governance, it's not going to work.

>> Yes.

>> It's a non-starter at this point.

>> Absolutely.

>> Especially with the scale and everything else. Let's talk about no-code and low-code in the data pipeline support of Databricks and end user data integration and transformation?

>> Absolutely.

>> What can you tell me about it? How do you help that?

>> Talk about the relationship with Databricks. Obviously you guys are on the ecosystem, Partner of The Year for data integration.

>> Yep.

>> BFFs.

>> Consistency is critical.

>> Good timing.

>> Are there any customer use cases that get you excited personally? As much as I'm sure they get you excited as a business, but just as a human?

>> I think that's .

>> What's next in this evolution for Databricks, you think? You're starting to see the openness coming. What do you see them doing next?

>> I love the serverless.

>> Yes.

>> Yeah.

>> It reminds me of taking some friction out of the managing all these clusters. I fetched some data on it, Savannah, and the openness too-

>> Yeah, it was a big announcement.

>> It was a big deal. Yeah, it was.

>> We have a little view into that and-

>> Oh, that must be nice. All right, well, on that note, what's coming next then? What can we expect?

>> Tomorrow.

>> What can you tell us next? What can you tell us next? What's the future for Informatica, first? We'll go that direction so you don't have to be too naughty yet.

>> So I'm not going to name a specific technology necessarily.

>> You can name what's better.

>> Okay. All right. All right.

>> I'll let it extend out. What do you hope to be able to say when we have you on the show at Databricks Summit next year that you can't say yet today?

>> I hear that so much. So you're going to bring a customer up here-

>> Sure. Absolutely. Absolutely.

>> Next time and we can tell the story together?

>> Yep.

>> I think that would be absolutely fantastic. All right, well, this has been thrilling. John, thank you for your insights, as always.

>> Appreciate it.

>> And thank you for coming to hang out with us, Rick.

>> Always a pleasure.