We've been using AI for years now. But at this point, it's hard not to wonder where it's actually making a difference (and chances are, your leadership team is wondering the same thing, too). So what does that next chapter of AI look like? Well, it's less about the models themselves and more about the connections between them.
Over the past few years, we've learned a lot from our experimentation with GenAI, and those lessons are shaping how we think about working with AI agents today.
One of the biggest takeaways from our interactions with AI is that success starts with having your data in order. No matter how sophisticated the technology is, it's hard to get meaningful results if the underlying data is fragmented or incomplete.
Another thing we’ve learned is that while the industry is trending back toward more unified platforms, most organizations still rely on a mix of different tools, each with a specific purpose.
The way we interact with software itself is even starting to change. For years, we've adapted our work to fit the applications we use.
Now, we see AI becoming the interface that helps us navigate those systems instead of requiring us to jump between them. This only works, though, if AI can access the right information and understand the context behind it.
But can it?
That's one of the reasons many leaders have been paying closer attention to the Model Context Protocol (MCP). Think of it as a common language that lets AI agents connect to different applications, data sources, and business systems more consistently.
Instead of building custom integrations for every tool, MCP gives AI a standard way to find information, understand context, and take action across the technology businesses already use.
For marketers, that could mean an AI agent pulling together campaign performance and customer insights to determine the right context at the right moment, without constantly switching between platforms.
For merchandisers, an AI agent can connect catalog data, inventory levels, search trends, customer behavior, and merchandising rules without requiring teams to manually piece everything together.
Ultimately, AI doesn't become more valuable simply because the models get smarter. Its real value comes from its ability to work seamlessly with the systems, data, and technology organizations already have in place.
This or that
Morning deep work or late-night productivity? Late-night productivity. I’m a night owl, and after I’ve had all my morning meetings and lunch (and maybe a workout), I like to get into my deepest creative work.
Light mode or dark mode? Dark mode. Whenever you work closely with development, switching to dark mode feels like a total inevitability at a certain point. It’s just so much easier on the eyeballs.
AI for brainstorming or AI for editing? AI for brainstorming. I like using voice mode to completely brain dump what I’m thinking and use it for a good back-and-forth. Plus, I already have the best editor out there — shoutout to The Edge’s own Michael Lee!
Who’s shaping agentic AI?
If you picture the typical AI builder, you're probably imagining a developer.
But the results from Bloomreach's Loomi Connect Hackathon last month told us a different story. Marketers, merchandisers, agencies, students, and AI-native companies all showed up ready to build, proving that agentic AI is attracting a broad audience of users.
In his latest LinkedIn article, Andrew Kumar, our Head of Loomi Connect, unpacks the trends behind the data and what they reveal about where agentic AI is headed next.
However you're approaching AI this year, I hope you keep experimenting and stay curious. Thanks for spending a few minutes with me, and I'll see you in the next edition of The Edge!
Bloomreach, Inc. | 700 E. El Camino Real, Suite 130 | Mountain View, CA 94041
If you do not want to receive our marketing emails, please unsubscribe here.
