# The Future Is Well-Architected: Keynote Highlights Auto-transcribed by https://aliceapp.ai on Thursday, 19 Sep 2024. Synced media and text playback available on this page: https://aliceapp.ai/recordings/3zo1D2BFGALhvpMYwWRVrtmgi06cqsK7 * Words : 1,233 * Duration : 00:06:43 * Recorded on : Unknown date * Uploaded on : 2024-09-19 20:57:41 UTC * At : Unknown location * Using : Uploaded to aliceapp.ai ## Speakers: * Speaker A - 54.99% * Speaker B - 13.87% * Speaker C - 12.0% * Speaker D - 19.14% ---------------------------- Speaker A [00:00:00] Good morning, good afternoon, and good evening. Salesforce architects, aspiring, uh, architects, and other trailblazers, the future is well architected. That's not just the title of the keynote. We believe that despite all of the incredible new technologies that you've seen here at Dreamforce, the fundamentals of good architecture remain. It's critical that we as architects, build solutions that are trusted, easy, and adaptable. And, uh, I don't know about this group, but I am so energized after seeing that great keynote featuring Zane and so many other talented architects, we honestly didn't want it to end. Um, and the good news is that here on Salesforce plus, it doesn't have to. I'm actually joined by three of the talented architects fresh from the keynote that you saw, asking questions to that talented panel at the end, Shelly Sousa, Chris Valentin, and Gillian Reynolds are here for some post keynote analysis. Okay. I started with the pillars of well architected, and it always has to start with trust. And, Shelley, you spend your days thinking about security in CRM, and now we had that great announcement, uh, about data cloud security, and we're starting to understand how they're going to differ a little bit. What's on your mind? How are you processing these changes we just learned about? Speaker B [00:01:25] Yes, Susanna made a great comment. Uh, trust, but verify. That really settled with me, and I think, uh, seeing all of these new tools in data cloud that we're able to utilize as architects is really, really exciting. Um, the first is the key management storage. Now, we have the ability to bring our own keys, uh, which is excellent for any of us who happen to use aws or other tools where we have external keys. We also have the ability to control our data with data masking and with role based and attribute based access controls. So that gives us the ability to sort of pull some levers back and forth when we want to create control models that are hybrid based. And it gives us also the chance to make sure that when our data is leaving our orgs, that we have a little more control over it. So it's sort of like having an adult in the room, so to speak, from a data cloud perspective. Speaker A [00:02:07] Yeah. Yeah, absolutely. Well, I think as architects, we often feel like we need to be the adults in the room. So this is good that Salesforce is equipping us with these technologies to build that trust among adults and the rest of our stakeholders. So, um, shifting over to the easy pillar. We, um, want to build solutions for our stakeholders that are easy to use, and it's hard to go anywhere here at Dreamforce without hearing about AI, and people are getting excited. We're talking about agents, um, a bunch of different stuff around this. So, Chris, one of the things we saw in the keynote was we'll now have the ability to merge, so to speak, our LLM data with our customer data and keep it proprietary. You asked a great question about retrievers in, um, data cloud. Tell us more. How are you going to bring this back to your customers thinking about retrievers? Speaker C [00:02:55] Well, I think with AI we have to realize that we're all prompt engineers, and with retrieval augmented generation, that's really going to take those prompts to the next level. We're going to be able to bring in structured and unstructured data from Salesforce into those prompts so that we can get the best possible result. And that's, I think, what we're all looking for. One thing I really like with the retrievers, we get to build them ourselves on data cloud, and then when we use those in our prompt in prompt builder, we get to see or get a glimpse of what that prompt is going to look like. And hopefully that's going to help to connect that prompt that we put in to the answer that, uh, Einstein gives us. So hopefully we can use that to get the best results possible from our prompts. Speaker A [00:03:49] Yeah, absolutely. And this is, I can't stress this enough, this is so much better than just typing questions into an LLM randomly. We've got to go through that trust layer. We've got to use our customer data as reference. This is absolutely a big advantage for Salesforce. Yeah. All right. And finally, adaptable. Uh, Jillian, our solutions need to be adaptable and evolve with business. We, uh, did see an announcement about something that architects have been really excited about and asking a lot of questions about data cloud sandboxes. So you work with all types of clients. How does the announcement of data cloud sandboxes change your thinking about release management? Speaker D [00:04:25] Yeah, well, those data cloud sandbox is definitely highly anticipated, and I would say in addition to that, the deployment tooling that we also saw highlighted in the demo and the keynote. So these really work together. And for me, it's all about thinking about application lifecycle management processes and designing those so that data cloud and our CRM platform changes can really be handled consistently. So that's going to include everything from actually deploying changes, moving our solutions between environments, sandboxes, all the way to production, and being able to treat those as one unit and deploy them in similar ways using tools and processes. And the other thing that it really unlocks which uh, is great is the ability to tailor our environment strategy and testing strategy when data cloud is included. So having those dev testing UAT sandboxes, right, that whole stream of environments and being able to build our CRM solution in data in the same place, move it together, test it, and the other thing I'll add is that ability to tailor which data sources we connect data cloud to in each environment is really important as that can often be very complex. Um, and so that gives us the ability to tailor that and be much more flexible and robust. So really for me it's all about consistency when it comes to application lifecycle management between data cloud and CRM. Speaker A [00:05:43] Yeah, I love that flexibility because when you're first setting up the sandbox, you may not need to bring in all those data sources. Um, you may not need to put that all in. You can just choose, choose and pick the ones that you want. All right, awesome. Thanks Jillian for helping us be adaptable. That's about all the time we have today to geek out on the architect keynote, but please bookmark this link on Salesforce. Plus there's a lot of good information. I feel like you're going to want to come back and watch this keynote two or three times. Okay, much more information for architects. We have a full website dedicated to our content. You can see the QR code on the screen. Architect dot salesforce.com again, go there and you will find things, um, about our framework as well as our diagram guides. And you'll also find decision guides on how you can make some of those tricky decisions in your Salesforce architecture. So go out, be trusted, keep it easy and stay adaptable. Thank you Shelly, Krist and Jillian. And thank you Dreamforce.