Navigating the AI Era with Skai’s Celeste AI: An Interview with Channel Baker’s Aram Howard

Artificial intelligence is changing how agencies work, and tools like Skai’s Celeste AI—the first GenAI marketing agent purpose-built for commerce media— are showing what’s possible when AI is used well. Instead of just automating tasks, AI is now helping teams think differently about strategy, find new opportunities in data, and move faster from idea to execution. For agencies focused on retail media, that shift is becoming a competitive edge.

How you approach AI makes a difference. Some teams rush to apply it everywhere without much direction, while others take a slower path, layering it into workflows with clear intent. The strongest results tend to come from teams that start with specific problems, teach people how to think with AI, and build it into their daily process in a practical way.

That’s why we sat down with Aram Howard, Account Executive at Channel Bakers. His team has taken a measured but effective approach to using AI, blending tools like Celeste into their work to reduce friction and uncover insights they might not have found otherwise. His perspective shows how thoughtful adoption can make AI feel less like a trend and more like a natural part of agency life.

Interview with Aram Howard

Aram Howard is an Account Executive at Channel Bakers, where he focuses on Amazon Advertising and data-driven strategy. 

With a background in analytics, automation, and marketing, he combines technical skills with a practical understanding of what drives results. A graduate of the University of Alabama with a degree in Marketing and a minor in Mathematics, Aram is known for streamlining processes, improving collaboration, and helping his team use AI to work smarter.

He is also a power user of Celeste, applying it daily to streamline workflows and surface deeper insights for his clients.

You’ve mentioned that Channel Bakers has taken a really intentional path with how the agency is using AI. Can you walk us through that?

At Channel Bakers, we’ve always believed technology should enhance the work we do, not replace it. That philosophy has shaped our entire approach to AI, Celeste included.

Rather than jumping on trends or using AI just to impress clients with buzzwords, we started by pinpointing real pain points in our day-to-day work. Campaign planning, competitive analysis, workflow automation, etc. Each of these areas presented opportunities where AI could add value. From there, we focused on solving specific problems with specific tools.

I’ve spent a lot of time training my team not only on how to use these tools, but also on when to use them, how to think alongside them, and just as importantly, when not to. That shift in mindset, treating AI as a true tool in our toolbox instead of another shiny dashboard, has made all the difference.

When you’re showing someone new how to think about Celeste, what do you tell them to help them wrap their heads around how they should be approaching it — even before they type that first prompt?

I usually tell people to think of Celeste less like a search engine or reporting tool and more like another strategist on the team. You wouldn’t walk up to a coworker and say, “Give me the top 10 campaigns,” and expect anything meaningful in return.

Instead, take a step back and ask yourself: What problem am I really trying to solve? What will the answer to this question help me accomplish? Celeste is most powerful when it’s part of your thinking process early on, not just when you need an answer at the end.

I encourage new users to start with either open-ended questions, hypotheses, or highly specific prompts. For example: “What are optimization opportunities for Sponsored Brands in this category?” or “Which keywords performed best in X category during Y time, with spend over Z, that ran out of budget early in the day?” This is where Celeste shines by contextualizing data, surfacing insights you didn’t think to look for, or crunching information at a scale no one could manage manually.

How has Celeste changed the way you work, and what’s surprised you about it?

Celeste has completely reshaped my expectations for how fast I can move from a question to a clear insight. Tasks that used to take hours, like digging through reports, aggregating data across ASINs, and researching keyword trends, now happen in minutes.

It’s especially useful during QBRs or post-tentpole reporting, when we’re analyzing massive data sets under tight timelines. What surprised me most, though, is how seamlessly it fits into our workflow. There’s no time lost pivoting data correctly or juggling multiple reports. You just ask the question, get your answer, and keep going.

I’ve noticed a real difference in client meetings, too. I can test hypotheses in advance and come prepared with multiple angles of insight. And if a question comes up unexpectedly mid-meeting? I can prompt Celeste on the spot and have a data-backed response in moments. That level of agility is a game-changer.

Where have you or the team taken an insight or recommendation from Celeste, made a change, and seen some initial results?

A great example came from a QBR with a consumer electronics client. We were peppering Celeste with questions, and one query flagged our Rewarded Ad performance. It was strong—especially during tentpole events—but when we dug deeper, we noticed redemption rates were surprisingly low.

That meant our assumed cost per conversion, which we thought was around $30–$50, was actually pennies. Armed with that insight, we doubled down on Rewarded Ads and saw performance nearly double compared to other campaign types in the category.

It’s now a core part of our strategy and will be a major focus for Q3. More than anything, it reinforced how Celeste doesn’t just spit out direct recommendations; it accelerates the process of asking smarter questions and uncovering insights we might never have stumbled on otherwise.

What aspect of Celeste do you wish you could apply to your non-professional life?

Without question, I’d love for Celeste to manage my day the way it manages campaign data—especially my to-do list. Being able to ask, “What’s the most efficient way to run my errands tonight?” and get a step-by-step plan would be incredible.

I’d also geek out over personal stats. I’m a data nerd, so things like “How many times have I blinked today? How does that compare to average? and How does it line up with my sleep schedule?” would be fascinating.

We’ll probably get there someday. For now, I’m more than happy having a faster, easier way to get to the answers I need at work.

Conclusion: Adopt AI with purpose

AI is reshaping agency workflows and client strategies, and Aram’s approach shows the value of being deliberate. By starting with clear problems, teaching teams how to think alongside AI, and focusing on practical use cases, Channel Bakers has turned AI into a real driver of efficiency and insight rather than a distraction.

For marketers, the takeaway is clear: AI works best when it’s built into the way you think, not just the way you execute. Treat it as a partner early in the process, use it to test ideas, and focus on the areas where it can truly speed up decisions or uncover opportunities. The teams that do this well will be the ones who see the biggest impact.

If you’re ready to see what this looks like in practice, explore Celeste AI from Skai. It’s designed to help marketers connect data, answer complex questions quickly, and uncover insights that drive results, all in one place.

For information or to see Celeste firsthand, please schedule a brief demo with our team.