Databricks & Microsoft Extend Partnership Through 2030s: Azure Agents

0:00Uh we've had a deep partnership with Microsoft for many many years and I'm super excited uh to play uh a video that I just recorded with a CEO of Microsoft Satanella and just after we spoke just a couple of days ago we actually closed uh a deepening of our partnership. Uh we're extending the partnership that we have well into the 2030s. Uh and we're super

0:23committed to Microsoft Azure Databricks. Satya and Microsoft have been great partners of ours and we're looking forward to deepening that further. Let's play the video. No, thank you so much for uh having me Ali again. Congratulations to you and the Databricks team. It's just fantastic to see what we got started on

0:45whatever now close to seven I mean eight nine years ago and where it is and it's just congrats to all for all the success you guys have had. Well, a lot of it is thanks to the partnership that we have between the two companies. Uh I wanted to ask you um you're actually at the conference uh two years ago here at data and AI summit uh and you know AI was just taking off and uh you know it's

1:07been crazy two years and actually you had some pretty spot-on predictions back then. I'm cu curious anything that's surprising to you that you've seen in the last two years and when you look forward uh how do you think the next two years going to evolve in this space? Yeah, I mean it's it's to your point, it's just been crazy in terms of the pace, right, which is and and like all tech uh changes. I think some things we

1:29always forget is it's the compounding effects of the all the scurves, right? There was the pre-training scurve and then there is now the test time compute scurve. Uh there is just the application tier getting more mature scurve, right? We've all learned even how to build stateful applications using AI. Uh and so I think it's that combination that's

1:52been pretty stunning to see and also the other thing is the evolution right Ali which is we started by saying wow I can now have chat intelligence that I can chat with then we said oh we can assign tasks to the intelligence and now we effectively can have you know digital co-workers right so we which work autonomously we're still in the loop but

2:14all three of this right when I look at even my usage of say something like co-pilot on a daily experience. All three modalities or form factors are part of my daily experience. And so I think that that's exciting to see uh the rate of progress, the compounding effects of S-curves. And quite frankly, at the end of the day, the enterprises and the developers being able to build on top of the platforms you're creating,

2:36we are creating. I want to turn to helping these enterprises actually get that value out of those autonomous systems. Um you know, we've been working together now. partnership's been eight years as you mentioned uh is when we started Microsoft Azure Databricks partnership together uh and over the years we've been integrating with the whole Azure ecosystem so Azure Databricks is truly native part of all of

2:59the Azure ecosystem you know over the years we added foundry uh power platform the whole SAP partnership we did is actually built on top of Microsoft Azure Databricks uh so I'm curious looking forward uh how can we create even more value together and how are you seeing the partnership evolving how Can we accelerate this together? Yeah, I mean that is the most exciting part. I mean I think it was pretty unconventional when we came together back then to say hey

3:23we're going to make this a firstparty service and so that's why whenever I think about uh Databricks I think of it as a firstparty service and it's just great to see that seamlessness of integration and as you said now we're connecting the graph even more right even what we've done with foundry and Databricks that means you can build these stateful multi- aent automations on top of Databricks uh with foundry uh

3:47or what we're doing with the power platform connector. It's awesome for anyone who's low code no code to be able to build applications on top of Databricks. Uh what you have done with SAP I think is a gamecher. So for us to be able to even between SAP Databricks and sort of all of these connectors we have the amount of value we can add to enterprises is I think you know uh you know pretty stunning. So I feel uh that

4:09what we've built is pretty unique. the fact that every day we get get up and as two organizations that are connecting the graph even more ultimately to add value to sort of enterprise IT enterprise developers and enterprises get more value out of what is the hard work they've done right one of the things I always think about Ali is if the new technology somehow builds on the

4:31investments they've already made and compounds the effects for them then that is a true value ad and that's I think one of the goals that both of us share and it's great to see us keep making progress on it. Yeah, 100%. Uh now, you know, there's more talk about AI than ever before. Uh you know, from you know, CEOs all the way down to practitioners. Um you know, there's a lot of hype. Uh

4:54but there's a lot a lot of value that's actually being created. Um you know, I'm curious from your vantage point, where are customers getting actual real value? Where can we separate the hype from reality? And you know, what's this going to look like? you know what's what's what are the things they should focus on if they really want to extract value from AI. Yeah. I mean I think just like how when somebody sort of said you know

5:16uh let's make an enterprisewide platform decision around analytics with Databricks right that's when you know that enterprises are all scaling beyond like a p here or there to making a bit of a platform choice so that they can have many use cases that can be built using a common technology base a common

5:38architectural pattern and that's where we are now right when I look at even between Databricks and foundry that's what we offer or Databricks foundry and copilot studio or power platform that's what we will offer so to me it's the top- down scenarios whether it's customer service marketing finance what have you I think that those are happening the more exciting thing even

6:00for me internally even at Microsoft is to see where I now have people going and taking because it's become standard issue right I mean it's not just GitHub copilot for software engineers But you know with something like copilot plus you know my ability to have a connector to Databricks to be able to build an agent that then I can automate some supply chain process you are now getting to a place where it's become standard

6:23issue where every employee inside the enterprise is able to you know either manage agents create agents orchestrate agents uh in order to create that next level of automation around them change the work the work artifact and the workflow low right that to me means that the compounding effects really show up

6:45the productivity gains because the ultimate you know measure for me always is not some AI benchmark uh it is the GDP growth in the real world that means it's the growth in a firm and that's I think where uh we're seeing absolutely real evidence that this is scaling yeah 100% that's well put you know getting these agents to really seamlessly integrate with that whole graph like

7:08extending that graph which we look forward to doing with Azure for many years to come and getting those agents to understand the data that they have inside the enterprise, the data intelligence. So, super excited about the partnership. Satya, thanks for taking time. Um, maybe we'll talk here again in two years. Absolutely. Thank you so much, Ali. Thank you.