1. There are so many places where commerce and retail media agencies can be experimenting with and adopting AI agents. Why is Xnurta focusing on reporting?

Reporting has always been a bottleneck for agencies. It’s low-leverage work that drains high-value people. At Xnurta, we analyzed over 20,000 client profiles and realized something striking: reporting doesn’t just make agencies less efficient, it limits growth.

We found that the average agency was spending 2-5 hours per client per month on reporting. If you’re an agency with 50 clients, that might mean that you’re spending 250 hours a month simply compiling reports. In the context of Amazon, that can mean taking dozens of static screenshots from seller central, the ads console, and AMC, stitching them together into a report template, and then weaving those slides into an analytic narrative.

That’s time you’re not spending on creative strategy, experimentation, or relationship-building. With our AI Reporting Agent, we’re not just speeding up reporting, we’re changing the function it plays. You give it a prompt, it generates a full client-ready report in minutes, in your format, using your data. Then, what could have taken 5 hours takes 10 minutes to compile and 20 minutes to analyze.

That’s a 90% time savings.

Think about how many more clients that agency could service if that 250 hours turned into 25!

2. What does “agentic AI” mean in this context?

A lot of solutions on the market today that use the term “agentic” are simply generative AI chatbots. They can analyze the context you give them and generate answers, but they aren’t independent actors.

This is the evolution: dashboards were passive, then AI copilots became assistive. What we’re building now are autonomous agents, AI that understands goals, takes action, and adapts in real-time.

In Xnurta’s case, our AI Reporting Agent doesn’t just analyze the data. It builds the report for you by determining what matters most to your client, formatting the output accordingly, and presenting it in a way that invites interpretation. Once you’re able to get your prompting and templates dialed in, you pretty much never have to think about report generation again: you have a new team member who does it all for you.

You’ll still want to review your reports and do your own analysis, but even that can be assisted. On one hand, you can give it SOPs, playbooks, and past reports to train it to proactively flag key callouts in your presentation. On the other hand, if you’re struggling to understand a trend or datapoint, it already has the full context of your brand’s advertising performance, allowing it to suggest potential causes that you might not see.

3. How are agencies reacting to this shift? Isn’t there some fear of being replaced?

Agencies aren’t afraid of automation; they’re afraid of not implementing it quickly or well enough. Humans will always be in the loop, but agencies know their edge isn’t in building pivot tables, it’s in synthesizing insights, pitching bold ideas, and navigating complexity. Our focus is giving our agency and brand partners the ability to do more with less. This doesn’t replace any single job, it takes a low-leverage (yet important) task off of the vast majority of your employees’ shoulders.

4. What type of insights does your tool offer today?

Right now, it’s focused on Amazon Ads and retail media metrics, but that’s just the beginning. Soon we’ll be layering in more insights from Amazon Marketing Cloud and Amazon’s DSP, and our roadmap also includes bringing in other retailers such as Walmart. The goal is to give agencies the ability to generate unified cross-platform reports.

We’re also working on integrating performance recommendations directly into the reporting layer, so instead of just seeing what happened, you’ll know what to do next. And long term, I see these agents becoming proactive—delivering alerts, insights, and even media plans before you ask.

5. What's your broader prediction for AI in retail media by 2027?

By 2027, agentic AI will be table stakes. Every serious agency and brand will be running media with intelligent agents acting across bidding, reporting, forecasting, and creative optimization.

Here’s where I think we’re headed:

● The rise of multi-agentic teams: At Xnurta, our vision is to give agencies and brands access to a team of agents that can empower individuals to get more done. We’re starting with a single reporting agent, but our plan is to go much further.

● Agentic AI will drive a new class of microbrands: AI will make operating lean teams easier than ever. You can launch targeted SKUs with minimal overhead. You can test and iterate quickly without large teams, updating messaging, A+ content, and even packing in real time in accordance with customer feedback. AI will democratize brand building just as Shopify and Amazon democratized selling online.

● Agents as co-pilots rather than a closed cockpit: The key will be trust. AI agents can’t be black boxes. We’re building tools that are transparent, reliable, and adaptable. AI needs to be able to explain the actions it takes, and over-communicate that to marketers along the way, keeping the human in the loop. We’re seeing Amazon take this approach themselves, requiring their GenAI tools to receive human input before making changes, building shopping assistants like Rufus that aren’t black boxes for brands, and enhancing measurement to give brands more insights into customer paths.

The future is bright. I think we’ll see a rise in fun and inspiring brands as agencies and the brands themselves become freed to focus on creative and strategic work.

Get in touch with our MarTech Experts.