Agentblazers: Shaping the Future of AI with the Salesforce Community

Hi, everyone. Welcome. I'm Meredith Brown. I lead up trailhead in the community here at Salesforce. We're really excited you're joining us for this special session about Agent Blazers. This is part of our week of AgentForce, and I'm absolutely thrilled to be joined by one of our top agent Blazers, Melissa Hill, dies, and we're going to talk about how the community is shaping the future of AI. So let's dig in. Hi, Melissa, thanks for joining us.

Hi, Meredith. Thanks so much for having me here.

All right. As you know, we recently introduced a new term for our trailblazers. That's our customers, our employees, our partners, and these are the folks who are gaining agentic AI skills. We call them Agent Blazers. And so I'm curious, what do you think that path typically looks like? You've been a trailblazer for a long time. What does the path look like to become an Agent Blazer?

I really think being an Agent Blazer is all about all the combined skills that I've already acquired through Trailhead, through experience, through the collaboration with the community, and then adding those agentic skills and the AI skills on top of it, and in my case, to really improve user experience.

That's great. So when you think about other folks who are becoming Agent Blazers, like, what skills are you working on right now? Is there anything that you found in Trailhead that really caught your eye?

So I love Data Cloud. That's an exciting part of the entire agent ecosystem that is probably undersung in my estimation. Being able to have all the data you need and have it actually together in one place so that you can access it is really the food that AI needs. And so I've been learning as much as possible about Data Cloud and learning how to access and leverage it.

Did you enjoy some of the trails in Trailhead where you're able to get hands on with Data Cloud?

Yes, yes. I love it and have actually gotten to implement it for a couple of folks already. So I love the hands on learning in Trailhead. I'm a very kinetic learner. Personally. I don't want to watch a video, I don't want to listen to audio. I want to do it and make my mistakes and learn from them.

Oh, absolutely. I made my own mistakes going through Agent Blazer, champion innovator. That's. That's definitely how we learn. So how do we take all the things that you've learned that I've learned, and how do we help other people take that knowledge and share this with the community?

So my very favorite Thing that that Salesforce has done already was the hackathon at tdx. It was an exciting opportunity to do a couple of things. I recru three of my girlfriends, my female friends, because I really believe that women collaborate incredibly well. And I think that's what AI is really all about. And the four of us came together. We don't work together on a regular basis. One of us is a developer, one an architect, one the flow queen. And I was the only one that had ever built an agent. But we collaborated. We spent little less than 16 hours together and built a working agent that actually can schedule our community event speakers for us.

Oh, that's so cool.

It was fantastic. It was the best learning experience. I was convinced that I was absolutely not going to use code because I don't write it anymore. And as it turned out, the time constraints and what we needed it to do, code was absolutely the best way to go. And so Melissa Hansen that was on the team, whipped out some code right quickly and we put it together and it's a wonderful agent. We can't wait to get that on Agent Exchange and share it with our fellow community event organizers.

That's awesome. I love that you found like a real need within the community. What other use cases are you thinking about? What kind of common challenges can we solve?

Sure. So I think one of the very first agents that I built I did in about an afternoon. And I've worked with an ISV partner for eight years and we had over 15 years worth of screen steps, information. Right. So release notes and how tos and all these different things that nobody's going to sit down and read. And it was such, such a simple process to be able to download that into a PDF format, put it in an Einstein data library, and then create an external agent on the application to be able to say, how do I create a volunteer opportunity? And it would give you step 1, 2, 3, 4, instead of what I usually do, you know. 57 tabs later, I still haven't found the answer I need.

Yeah, I mean, I was thinking about that for Trailhead as well. I mean, our learning agent, all the things that we could do, we were like, I need an implantation guide. Give me the step by step. So I think there's just so many different, different things that, you know, I'm thinking about as well as we, we help people skill up on Salesforce products and being able to bring that agent into Trailhead as well.

My favorite thing may be that I don't have to write code. I probably have Failed platform developer one more times than anybody else in the community and haven't passed it yet. I'm determined to do it has made me mad now but, but to be able to do it, to use Agent Force and to use AI without having to write code, you know, I'm learning and scaling up more on the large language models and the external models that you can bring in and what that means from a security standpoint and the Einstein Trust layer and all the pieces and how they come together. Building the agent itself is the really simple, easy part. But then understanding all the rest of the architecture that fits around it to make it secure, to make it a good user experience, that's exciting to be, to be able to pull all of that together. And TRAIL has been great in helping learn that.

Yeah, I mean, I think what's been really resonating with me as I've talked to lots of folks in the community is like, if they're just getting started, they're like they can get the AI skills that they need. If they want to go in and build their first agent, they can do that. But it sounds like you've moved to the next stage where you're like, okay, I've got it built, now I need to figure out how to optimize it. So for the folks who are maybe earlier in their journey, what kind of advice might you have for them?

Just do it. That's don't be risk averse, that you can't make a mistake and if you do, it's okay. You'll learn more from the mistake than you will if you don't do anything at all.

I love that you say just like experiment. I couldn't agree more. Sometimes it just feels so overwhelming. You're like, I've never done this before. I mean, Agent Force didn't even exist a year ago. I think we all have to remember, like we're all learning together. Just get in there, just do it. You just learn so much. And I think what I love about the community is all of these events where people are coming and they're learning from each other. So that's just an amazing thing too. It's like you can find someone else in the community, be like, well, what was your journey? How did you do it? How did you apply these different use cases for your business? There's just so much we can learn from each other.

The most recent opportunity I had to do this, I was at the Open Source Commons community Sprint in London a couple of weeks ago and we. Someone brought up the idea of Using AI to help document their nonprofit cloud orgs. Right. And we were lucky enough. We had a CTA there, we had a developer there, people who were not in the nonprofit community as well as lots of nonprofit folks. And we actually got started on all the requirements and all the thought processes and the consequence scanning to build an agent that would help us document the nonprofit orgs rather than someone having to physically do that, you know, on a regular basis to help them. You know, most nonprofit orgs don't even have a full time salesforce administrator, so they need a little bit of additional assistance in keeping up with what's going on in their salesforce organization. Very collaborative effort.

Oh, I love that. So what other types of like successes or challenges do you think that we can, we can bring to the table to inspire more folks to just get in there and try it? Like you said, just do it.

So I really, really feel like that the user experience is so critical. Right. That's what I think. That's why the whole AI situation took off a couple of years ago. Nobody had to train you on how to use it. Now we as administrators and developers and architects, yes, obviously we need some training. But for the end user, that's always the difficult part. Right. And that wasn't difficult. With AI, anybody can type a question. You know, it's kind of like, who trained you on Slack? Nobody. You just go in there and start talking to people.

That's right.

And I feel like that AI can provide that same user experience that makes it so much simpler. That takes away, oh, you know, I don't use computers. My mother says that I've never used a computer. I'm like, mom, you have a smartphone. It's a computer. Right. But the interfaces that she uses are so simple, you know, texting back and forth. And that's what AI provides in the form of agent force is a conversation. You don't have to understand how it works. You don't have to put in the, the right question or the right answer. It just works. And so if that doesn't inspire you to see how you can make your end users lives simpler and easier, then I don't know what would inspire you.

Okay, so speaking of inspiration, what's next for you in your journey? What do you want to learn next?

Oh, well, so I really want to help people understand that AI is not a big scary hairy thing and it can truly automate things that, that should be automated. Right. I mean those tasks that on a regular basis, the one that I'm working on right now, is specifically for nonprofits. But at the end of the year, every year, my husband says, okay, do you have all your receipts from your tax donations? And I'm like, no. So he starts calling wherever I've donated and asking for the receipt. What if you could just go to the agent where you usually make the donation and say, can you give me a PDF of all the donations I've made this. And print it out and hand it to your husband. You know, so simple. Aggregate the data. The data is there in whoever you donate to, and you don't have to understand how spreadsheets work or flows, work, or you just say, give me a record of all my donations. I want to print it out.

Yeah, I'm excited about that. Like, getting some time back in your day, being able to be more creative. I feel like in some ways, it's like we can all just be a little bit more human, too, because some of the annoying things that we have to do, like the spreadsheets and the formatting and all those things kind of go away. And now we can work on the things that, like, make us more human and more creative.

So I think that's what I love your nonprofit staff. Time to do the things that AI can't do.

That's right.

That's right. AI can't give me a hug.

That's right. Yeah. The human aspects of it. Yeah. That's so great. Well, I just want to thank you so much for spending time with us today. I learned so much from you, your journey. It's super helpful. Thank you for your story about the community coming together at the events. I love that story. I hopefully this is an inspiration to everyone to go out and get their agent Blazer Champion. It's free. There's nothing really stopping you. So, you know, go out and do it. Thank you so much, Melissa. I really enjoyed spending time with you today.

Thank you, Meredith.