Salesforce Admins Podcast How Can Admins Use Slack to Manage AI Agents More Easily?
Today on the Salesforce Admins Podcast, we talk to Jim Ray, Director of Developer Relations and Advocacy at Slack.
Join us as we chat about enhancements to Workflow Builder, the Slack features everyone should be using, and the future of AI and Slack.
You should subscribe for the full episode, but here are a few takeaways from our conversation with Jim Ray.
Usability upgrades to core Slack featuresJim and his team have been improving the core Slack experience. “We’ve really been focusing on those little paper cuts, the smaller features that just never quite made it into the next release,” he says. So essentially, Slack is a little better everywhere.
One area they’ve focused on is Workflow Builder. In particular, they’ve added new branching functionality—allowing admins to create conditional paths like “if/then” logic.
They’ve also been hard at work upgrading canvases and lists. With new data like AI-generated meeting notes, canvases provide a central place for all the relevant information. Jim also urges you to check out lists as a sort of “semi-database” for data you want handy in Slack.
Slack’s AI vision centralizes agent interactionsFor Jim, Slack is one of the best tools to interact with and fully take advantage of the AI agents you build with Agentforce. It’s already the place your team communicates with each other, so why not be able to loop in AI teammates?
“In the same way that Slack is the single place where every person in your organization is communicating, now it's the place where you're all working with those agents,” Jim says. And so his team is looking at how Slack can bring together every AI agent your team uses, whether they’re built in Agentforce or another third-party platform.
Dreamforce 2025 will spotlight Salesforce-Slack integrationI asked Jim for a sneak preview of what he has in store for Dreamforce 2025, and he did not disappoint. We’ve come a long way with tools like Salesforce channels and the ability to deploy an Agentforce agent directly to Slack. Now his team is working on ways to deepen the integration. So look for improvements to search and embedded Salesforce data in the coming year.
There are so many more great insights into how you can get the most out of Slack in the Agentforce era, so be sure to listen to our full conversation with Jim. And be sure to subscribe to the Salesforce Admins Podcast so you never miss an episode.
Podcast swag
Learn more-
Salesforce Admins Podcast Episode: What Can Salesforce Admins Do with Slack Integrations?
Admin Trailblazers Group
Social
Full show transcript
Mike: Welcome to the Salesforce Admins Podcast. This week we're catching up with Jim Ray from Slack because it's very conversational. He's going to walk us through what's new in Slack from major releases that you may or may not have been paying attention to on Workflow Builder, Lists, and Canvases to, of course, all of the latest around AI and Agentforce. Plus, we're going to talk about what's coming at Dreamforce this year and how admins and developers like yourself can get the most out of Slack. So whether you're all in on automation or just exploring Slack's potential, I promise you there's something here for you. So you've already got those earbuds in? Let's get Jim on the podcast. So Jim, welcome back to the podcast.
Jim Ray: Mike, thanks so much for having me. It's a pleasure to be back. I'm excited to talk to admins and maybe we'll talk a little bit about what's coming at Dreamforce.
Mike: I mean all the things, because it's not that I didn't want you back sooner. It's just a really big platform and there's so many things to talk about.
Jim Ray: Of course.
Mike: But we use Slack every day at Salesforce and a lot of our customers do too. So it's been a while since we've chatted, but what's some of the big stuff that Slack has come out with this year that admins should be excited for if they haven't been paying attention?
Jim Ray: Oh, that's such a great question, and obviously we use Slack every day as well, and nobody uses Slack quite like Slack uses Slack.
Mike: Oh.
Jim Ray: And I think that's probably true for Salesforce in some ways too. But we've been really excited to talk to some of our admins. We're continuing to learn more about the Salesforce developers in particular, and so always excited to hear about the use cases there. A few exciting things that have come out recently that I think will be pretty interesting. So our Workflow Builder product, so this is our no-code automation product that's built right into Slack, I think that's probably something that a lot of Salesforce admins and Slack admins are using. I like to think of Workflow Builder as the front door to the platform. And it's certainly adjacent to the platform in a lot of ways. So we've had some really great development happening in Workflow Builder. Something I'm actually really proud of is we've been really focusing on the core user experience of Workflow Builder. We've got a whole team that's working on something that internally we call Back to Basics, which is really just all about focusing on those little paper cuts or the smaller features that just never quite made it into the next release. So they've been building that out. So hopefully just the experience of using Workflow Builder is a little bit nicer, which is something that we always aim for.
Mike: I never found it bad.
Jim Ray: It's not that it's been bad, it's just that as the-
Mike: To be fair.
Jim Ray: Yeah, it's a great point. As the surface area of Workflow Builder has grown, we've added lots of new functionality and we've just needed to go back and polish here and spit-shine there.
Mike: Sure. I mean, for me as an admin, it was the most intuitive thing that I could immediately jump into and kind of feel like I made a difference when we got Slack and when we started setting up channels. Because I think you probably hear this a lot, but to me the fallacy of, well, Slack is just where we communicate. No, you can actually guide the communication, and we do that with our team to help get the right information to the right spots. Because there's data and then there's contextual data, which is the conversation, and that's where we use a lot of workflows. And I was almost overwhelmed by the number of options that I had on ways and things I could do with the conversation or the input, like we use it, like an input form. I was like, "Wow."
Jim Ray: Right, exactly.
Mike: "Oh my God, now I got to go back and rethink this." It wasn't something where I was diving into it and thinking, "Okay, well, I'll just figure out what I got and just work with what I had." It was showing up to the grocery store and being like, "Oh, so you guys really do have everything? All right, cool." Kind of wasn't expecting that. I was expecting three kinds of cheeses and you got a whole cheese counter.
Jim Ray: That's really great to hear. And even with the addition of things like Lists and Canvases, which are some features that we're continuing to expand on, we released those a while ago and obviously we use them a ton internally, but we're hearing from our customers that they're putting more and more of their mission-critical data. Canvases are fantastic because it's a really lightweight form of documents. You're not overwhelmed with all the different formatting options, but they're still really quite powerful, still a great way to capture information and add context to your channels. And my team uses them with Notes and with some of the AI generated notes that are happening within huddles and things like that. But I think Lists are another place that we're seeing a ton of value with things like Workflow Builder. So you were saying you've got an input form, but where are you going to put that once somebody has filled out the form? Lists are a perfect place for that. And again, it's not a full-blown database or even as powerful as something like an Excel or a Google Sheet, but if you just need to capture that data and you want it native inside of Slack, something that's searchable inside of Slack, something that you can easily add workflows to, I love Lists. And so they're both really fantastic features.
Mike: I haven't used Lists. I can't speak for them yet, but I have used Canvases.
Jim Ray: If you don't need something as full-featured as Jira, for instance.
Mike: Maybe that's why.
Jim Ray: My team, we use Lists a lot to track the projects that we're working on. We don't need a full suite of Jira tools. We're not working [inaudible 00:06:19] and things like that. So it's actually a really great way to do some lightweight project management as well.
Mike: Ooh, I like that. You said a magic word, AI.
Jim Ray: Yes.
Mike: So I feel like Slack's built for AI because AI everywhere is just conversation, and I'm like, "Well, naturally, that's Slack." So talk to me about Agentforce and AI and Slack.
Jim Ray: Absolutely. Before I jump to AI, I do want to mention one more Workflow Builder feature-
Mike: Oh, yeah.
Jim Ray: ... that I want people to check out is this is probably the longest standing request, which is the ability to branch your workflows. Essentially think of them as an if statement or a case statement if you're a programmer. But this was, when we gave people the ability to start doing some automation, the first thing they said is, "Oh, I need to branch my automation." So if they hit this button, do this thing, and if they hit this button, do that thing. We have finally built that into the product. It's a really fantastic way to make your automations even more powerful. If you are a programmer or if you're a developer and you're building on the Slack platform, one thing that I like about the new branching feature is that you can actually handle a lot of the, for lack of a better word, validation that you would normally be doing on the back end. Now, you can do a lot of that on the front end inside of Workflow Builder itself. So quick plug for branches there, another fantastic feature that we rolled out to Workflow Builder. And it's going to make those workflows even more powerful, but now we're ready to talk AI.
Mike: Yeah, no, that's good. I do like that. I kind of want to play with that now.
Jim Ray: It's really great. So yeah, so with AI, obviously, and Mike, as you implied, Slack is the best place to interface with the agents that are popping up in every single application that we use. And in some ways this is a return to form for Slack and the Slack platform. So when Slack launched in 2014, part of the reason that we were able to be so successful is we were riding this wave of a few trends that were happening in the industry. One of them, I'm sure you remember, DevOps. And DevOps was hugely successful to Slack being successful because a lot of our earliest customers were startups, engineering-focused organizations, media organizations, places that understood kind of the basic premise of DevOps, and we could just plug Slack right into it. So many developers looked at Slack and they were like, "Oh, this will fit right into my GitHub pipeline or my Jenkins pipeline, and then I can just pipe messages right into Slack. I can have all my entire engineering team in a channel and they can all come take a look at this." But as the platform grew and evolved, we added all of these features for interactivity. We added Block Kit, which is our UI composition framework. And now that we have all of these large language model and other generative AI tools, it is kind of this return to form for the platform in many ways. And so 10 years ago, if you wanted to build a Slack app that interacted with you, it was almost like a command line more than anything. It wasn't a natural place to chat. You would kind of send off some commands and then wait for the response. It felt very much like using a terminal or a Unix command line or something like that. But now with all of these LLM tools that are being plugged into all of the various systems of record, Slack is of course the place to bring all of that into the same place where you're working with your colleagues. And that's kind of been the basic pitch of the platform for a long time.
Mike: So I do want to touch on that because it's a thing that I've been thinking of a lot. And of course it's the admin podcast, but that doesn't mean that admins don't code.
Jim Ray: Sure.
Mike: Lots of admins code and they just choose to identify as admins. Some developers only code. I don't want the perception to be, well, Slack is just for admins because they don't have to write code to do anything. But I think as you sit and look at when you're working with your team, whatever, of stakeholders, it's an admin and a developer, especially at a large organization, probably an architect. There's a lot of roles involved. Slack can really interface with a lot of different products that you have across the enterprise. What do we do for admins that had to be like, "No, let me talk to my developer because Slack is very developer friendly"? When we say Slack is very developer friendly, what do we mean by that?
Jim Ray: Well, I like to think that our platform, first and foremost, is just one that developers really like to use. Again, that was something that was really critical to the success of the platform pretty early on. It doesn't feel like developing on a piece of enterprise software. And so we would hear from these developers that were used to these massive Java stacks, or they had to use an SDK that they weren't familiar with, or even a programming language that they weren't familiar with. And the premise for the Slack platform from the very beginning, and this is still true today, is that you can bring whatever programming language you want to. We make some first party SDKs available, but you don't have to use those. We have some frameworks that are even a little bit higher level than the SDKs. Those are all available in Python, Java, and JavaScript, but you don't have to use them if you don't want to. We have some amazing community-developed SDKs and frameworks as well. So that's the initial thinking as to what makes developing on the Slack platform great. But then the other part is that you can do things with the Slack platform that really you couldn't do with other forms of communication. They just didn't make sense in the context of email, for example. Even if you wanted to use email as the ultimate endpoint, it kind of didn't make sense because an email inbox is a one-to-one relationship with every person in the organization. Well, there's no really good way to collaborate until you pull people into these channels. And so that was another part of Slack's big success we kind of created, and again, we were pulling from a lot of what was happening in the world at the time, but we pulled a lot of these ideas and we put them into this kind of channel-based interface. And now you bring groups of people together that are working on a similar project, similar feature. And then when you layer on top of that, the platform pieces, that's where it starts to get really powerful. So again, think about your engineering org within your organization. So you've got a team and they get an alert from somebody has pushed some code and the tests fail. Well, now everybody on that team can see why the tests failed. So the engineer who pushed the code, that test probably wouldn't have failed if they knew what the problem was. So another engineer that might have a different bit of context, oh, I just checked in this bit of code, that's probably why your tests are breaking, for example. They can see that. And so now you're maximizing that visibility. And so when we layer on these other tools, to bring us back to this AI, so now every tool that you're using in your organization, they're starting to add these agentic features. And we were talking about, so if you're an admin, well, Slack now becomes the place. Rather than having to administer an agent in every single platform that you use, you bring those agents into Slack, so now you have a single place to work with them. In the same place that Slack is the single place where every person in your organization is communicating, now it's the place where you're all working with those agents. And so we're not there yet. We're still working on it. Obviously Agentforce is the best place to, if you're just getting started with this and you have a ton of data inside of Salesforce, of course Agentforce is the place where you're going to go, maybe try out some of these agents built on top of your data. But as we're seeing many of our partners, many of the other SaaS tools that our customers are using, all of these agent features are being built on top of the systems of record that they manage. And so what we've got, we've made those same AI features available to everybody on the platform. So now anybody can build an agent that works inside of Slack. And if you're an admin, this is fantastic news because now you've just got one place to manage all of these.
Mike: And we use agents a lot at Salesforce. It's kind of fun. Sometimes I don't know where to look for my approvals.
Jim Ray: Yeah.
Mike: One thing to look forward to, bury the lead there, Dreamforce, it's coming now.
Jim Ray: Absolutely. Absolutely.
Mike: They just announced the band. Are you going to go?
Jim Ray: I have never made it. Usually, because I'm so swamped, I've never made it to see the show.
Mike: I know.
Jim Ray: But we've got two choices this year, I think. Is this the first time that there's going to be two different stages?
Mike: I don't know. I feel like they're doing that now because there's a generation of us that are coming up in the years, and then there's a younger generation. I remember a couple Dreamforces ago, I didn't even know one of the opening acts.
Jim Ray: Oh, boy.
Mike: And I was like, "Oh, I didn't know what my parents feel like now when I was a kid, and be like, "But Mom, of course it's Michael Jackson."
Jim Ray: Right, exactly.
Mike: "Duh." And that happened. I'm with you though. I have been once to the concert because that year we had the admin keynote the first day and we could go and it was almost like, I don't know, a celebration for us because it's like, "Yay, our keynote's over, we can go."
Jim Ray: Oh, nice.
Mike: And it was fun because it was Foo Fighters. I don't know if you were around then. I can't remember the year, but it was the one year it felt like it was out in a field or somewhere.
Jim Ray: I think that was our first year as an official Salesforce company.
Mike: Okay. I went, we got in, it was like, "Okay, I really need something to eat." They had hamburgers and stuff. I had that. And I was like, "Oh, the stage is that way. Let's walk that way." We got to what we thought the stage was, it was a giant TV screen.
Jim Ray: Oh, no.
Mike: Because they had all these satellite stages.
Jim Ray: Yeah, exactly.
Mike: And I was like, "Oh, well, this isn't so bad." And it was Foo Fighters, which was awesome. It was one of the years that Dave Grohl broke his foot, and so he was in that really cool throne. But I mean, of all the people that could still put on a show, he put on a show.
Jim Ray: Nice.
Mike: And I looked over and I was like, "Well, no, it looks like the stage is over there." Long story short, we ended up walking to all three of the satellite stages and never found the main stage because we kept thinking we were. And I was like, "You know what? I think this is the point at which I need to stop going to the concerts because-"
Jim Ray: Yeah, there you go. And so this year we have a choice between, and you were talking about that generational divide, so it'll be very obvious. We've got Metallica on one stage and then Benson Boone on the other.
Mike: Yes. Yes.
Jim Ray: I can definitely remember buying Metallica on cassette tape.
Mike: Yeah. Yep. I used to buy Metallica stuff when you could go to the convenience store and buy a glass bottle of 16-ounce Mountain Dew.
Jim Ray: Oh, wow. Now there's a throwback right there.
Mike: Everybody's like, "Oh man, old dudes on the podcast." Anyway.
Jim Ray: Yeah, exactly.
Mike: Let's give admins and developers some sneak peek stuff or what are some of the stuff, Jim, maybe that you're working on for Dreamforce this year?
Jim Ray: Well, everything is still very much in flight.
Mike: Oh, absolutely.
Jim Ray: We just had our half year-
Mike: Full-on forward-looking statement.
Jim Ray: Absolutely. We just had the half-year company kickoff, but it was super exciting. The things that we're most excited about, obviously continuing to deepen the integration between Slack and Salesforce. And I think certainly from the Slack perspective, but I think even from the Salesforce perspective, that's really going to be a big story at Dreamforce, is deepening that connection. We've had things like Salesforce channels and these agents, the ability to deploy an agent from Agentforce into Slack with just clicks without having to manage any of the deployment bits. So super excited to just see that deepening continue to happen. And we've been working behind the scenes on some things that I think will continue to make Salesforce even more embedded and useful with the context of Slack, a lot of them having to do with things like search. And so we've been releasing these new features to search that allow you to search not just your Slack corpus, but your Salesforce data and your Google Docs data and your Microsoft data. And so continuing to broaden and make search that much more useful. I feel like search is one of those hidden superpowers inside of Slack. People will sometimes come ask me a question and I don't know the answer, but I know how to search for it and where to find it.
Mike: Oh, literally how to search for it. Ha-ha. I see what you did there.
Jim Ray: Well, and you kind of become the oracle within an organization in some ways. But there are some things that we can do on the product side to make some of those things a bit more obvious. So I think that'll be good. I'm excited about our continued partnership with Heroku. Heroku is really doing some really amazing work. And they're in a big reboot phase right now over there in Heroku, and they're really leaning heavily into AI. But if you are doing custom development work, Heroku really is the best place to deploy those apps, especially if you want those apps to show up inside of Slack. But some of the work that they're doing with their custom inference and the databases that they've got for AI tools, I think is really, really cool. So if you're a developer, if you've got a team of developers that's really interested in pushing pretty hard into some of this custom AI tools, it's probably time to give another look to Heroku and see what they've got going on. Really, we're doing some workshops with that team as well, and so we'll be showcasing how to build great custom apps, deploy them to Heroku and have them show up inside of Slack.
Mike: Oh, very cool. Awesome. Well, thanks for making time out of your day, Jim. And I mean, you're always welcome back to the podcast. I'm sure after Dreamforce, we should have a debrief and be like, "Okay, remember all those things we talked about? Here's where we're at."
Jim Ray: Absolutely. And by then, we'll have figured out what we're actually going to be talking about. So that'll be great.
Mike: That's always the plan. That's always the plan. Jim, it's great to have you on the pod. You're always welcome back. So thanks for keeping us enlightened.
Jim Ray: Excellent. Thanks so much, Mike. I appreciate it.
Mike: Well, that was a Slack-tastic time with Jim. You know I had to throw that in, try something, right? I appreciate him giving us a sneak peek on what's ahead for Dreamforce. It's going to be exciting this year. And enhanced workflows, smarter agents, I'm telling you, there's just never been a better time for admins to explore what Slack can do. Now, if you enjoyed this episode, do me a favor, share it with a fellow Salesforce admin or somebody else in the community that you think would enjoy listening to it. And until next time, we'll see you in the cloud.