Chloe Maraina has spent her career at the intersection of data science and creative strategy, transforming massive datasets into narratives that drive business growth. As an expert in Business Intelligence, she has witnessed firsthand how the shift toward agentic AI is redefining the relationship between human intuition and machine efficiency. Today, we sit down with her to explore the release of Adobe’s CX Enterprise and what it means for the future of customer engagement. We discuss the profound evolution of the marketing professional’s role, the technical hurdles of embedding unwritten business wisdom into automated systems, and the intensifying rivalry between the giants of the enterprise software world.
The following conversation delves into how modern organizations are moving away from episodic project work toward a model of continuous, AI-driven interaction. We examine the mechanisms of Brand and Engagement Intelligence, the strategic integration of third-party AI models like those from OpenAI and Anthropic, and the shifting boundaries between marketing platforms and traditional CRM systems as they vie for control over the customer journey.
The shift from being an “artisan crafter” to an “editor” of AI agents represents a major cultural change for marketing teams. How do organizations practically manage this transition, and what does it mean for the speed at which a company can launch complex product variations?
This transition is truly about elevating the marketer’s role from manual execution to high-level orchestration, effectively giving everyone a promotion in terms of their strategic impact. Instead of spending hours drafting a single piece of content, employees are now trained to fine-tune agents and perform rigorous quality checks on the output generated by these systems. This shift moves the entire workflow from being episodic—where work happens in bursts—to being completely continuous, allowing for a relentless pace of production. By utilizing task-based agents, companies can suddenly handle a much higher volume of marketing capacity, which is essential when you need to push out thousands of product variations to meet diverse global demands. The sensory experience of a “launch day” changes from a frantic scramble to a smooth, monitored flow of data-driven assets hitting the market in real-time.
Brand Intelligence and Engagement Intelligence systems are designed to automate high-stakes decisions like “next-best actions” for loyal customers. How do we take the “tribal knowledge” that lives in people’s heads and turn it into code, and what are the dangers of letting probabilistic models handle these sensitive interactions?
One of the biggest challenges is that business context isn’t always neatly organized in a SharePoint folder; it’s often hidden in the “tribal knowledge” of long-tenured employees or buried deep within legacy applications. To codify this, we have to look at the policies and rules that have been built into our software over decades and use them to ground Large Language Models (LLMs). By layering small language models on top of a secure data layer, we can create a governing structure that ensures the output remains within brand standards. The risk of relying purely on probabilistic models—which essentially guess the next word or action—is that they can lose the plot of a brand’s specific identity. However, by marrying these probabilistic powers with deterministic business rules, we create a safety net that protects the customer experience from the “chaos” that unguided AI can sometimes introduce.
With the introduction of a shared skills catalog featuring native platform integrations from providers like OpenAI and Anthropic, the possibilities for customization are vast. What should be the primary criteria for selecting specific agent skills, and how is the underlying data kept secure?
When selecting skills from a catalog, the primary criteria must always be the specific goal of the workflow, whether that is deep data analysis or rapid content generation. Developers should look for agents that have access to “common capabilities” across the enterprise, ensuring that a skill used in one department can talk to a skill used in another without friction. In terms of security, the architecture must start with the data layer at the very bottom of the IT stack, acting as the foundation for everything else. This ensures that while we are using powerful external models like those from Nvidia or AWS, the sensitive customer information remains governed by the organization’s internal standards. It’s a delicate dance of opening up to the best technology available while keeping the “keys to the kingdom” behind a strictly managed orchestration layer.
We are seeing a significant “collision course” between marketing platforms and traditional CRM systems like Salesforce as they both move toward managing B2B buying teams. What are the strategic benefits of approaching customer management through a CX-driven lens, and how can companies handle the inevitable interoperability issues?
The strategic advantage of a CX-driven platform over a traditional CRM lies in its ability to manage the entire continuous customer journey rather than just a static sales record. By focusing on stewardship over customer engagement, platforms like Adobe are now building tools specifically for B2B buying teams, recognizing that a purchase decision is rarely made by a single person. This encroaches on the territory of giants like ServiceNow and Salesforce, who are also expanding their reach, creating a landscape where everyone is essentially in everyone else’s lane. To navigate this, businesses must prioritize interoperability, treating these vendors as “frenemies” who must work together within a single enterprise stack. Success in this environment requires a tech-buying calculus that values how well a platform integrates with others just as much as its standalone features.
What is your forecast for the future of AI-powered customer experience platforms?
I believe we are heading toward a future where the distinction between different enterprise “apps” will almost entirely disappear, replaced by a seamless, agentic orchestration layer. We will see a shift where the “heavy lifting” of embedding business rules into workflows is handled entirely by AI, allowing human editors to focus purely on the creative and emotional nuances of a brand. As these platforms become more adept at managing B2B buying teams and complex loyalty goals, the “chaos” we see in the market today will settle into a new standard of hyper-personalized, real-time engagement. Ultimately, the winners will be the organizations that can most quickly codify their unique human insights into these digital systems, turning “tribal knowledge” into a permanent, scalable competitive advantage.
