
Sign up to save your podcasts
Or
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.
Jim 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.
For 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.
I 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.
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
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:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
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:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
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:
Jim Ray:
Mike:
Jim Ray:
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 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:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
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:
Jim Ray:
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:
Jim Ray:
Mike:
Jim Ray:
Mike:
The post How Can Admins Use Slack To Manage AI Agents More Easily? appeared first on Salesforce Admins.
4.7
200200 ratings
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.
Jim 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.
For 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.
I 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.
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
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:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
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:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
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:
Jim Ray:
Mike:
Jim Ray:
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 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:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
Mike:
Jim Ray:
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:
Jim Ray:
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:
Jim Ray:
Mike:
Jim Ray:
Mike:
The post How Can Admins Use Slack To Manage AI Agents More Easily? appeared first on Salesforce Admins.
4,341 Listeners
1,167 Listeners
3,219 Listeners
424 Listeners
3,981 Listeners
3,797 Listeners
1,035 Listeners
5,980 Listeners
2,987 Listeners
5,443 Listeners
227 Listeners
25 Listeners
265 Listeners
1 Listeners
1,294 Listeners