Building better conversations: How Aeroflow Health leverages AI agents to evolve customer relationships

In 2025, we’re all well-accustomed to text messages from brands, but the interactions often feel one-sided. In response to “Your appointment is scheduled for 1:15pm on Tuesday, press C to confirm and X to cancel” there’s not much we can say besides “C” or “X.” It hardly seems like much of a conversation. So, why do texts feel so one-sided?

Marketers typically parse consumer responses with simple rules. If the customer needs to do anything more complicated than confirming a repurchase or appointment, they often need to switch to a different channel—for example opening an app or web browser to make changes to an order. Ask someone to switch from one platform to another and brands create friction, which leads to customers losing interest.

Using AI agents to evolve text conversations

Aeroflow Health is a healthcare-focused consumer goods provider with a Mom & Baby division that helps pregnant and nursing mothers find medical devices, such as breastpumps, that are covered by their insurance. They use Braze for an SMS “re-ordering” journey—customers who are ready for a repurchase receive an SMS asking if they’d like to reorder in the same size. If the answer is yes, the customer has a great experience: she sends a simple text and gets the items she needs. But if the answer is no, the customer’s journey is more complicated: she needs to go to the website to make a new purchase.

Areoflow has started using the BrazeAI Agent Console™, now in Beta, to automate this journey. Now if a customer says she’d like a new size, an AI agent responds to ask what size she’d like. Unlike a traditional rules-based journey, the agent is able to make sense of the customer's natural-language response: first to infer and confirm that a size change has been requested, and then to interpret the customer’s description of the size she’d like. From there the agent takes the customer’s answer—e.g., “next size up please”—and tries to find a matching product in Aeroflow's catalog. If the customer confirms, the agent places the order.

Removing friction and easing the order process

In essence, Aeroflow is using AI agents to replicate at scale what any good (human) customer service professional would do: respond to customer feedback in natural language, confer with a product catalog, and then confirm and place an order. Simple as this seems, traditional rules-based systems have never “cracked the code” of providing this kind of frictionless experience. Crucially, Aeroflow’s customer doesn’t feel like she is talking to a chatbot; she’s simply talking to a brand that understands her needs and responds to them in real time. Best of all, the customer doesn’t have to change channels—she can confirm her purchase right in the conversation.

“Our customer used to have to reach out to us to change size,” Brian Munn, Senior Marketing Automation Manager at Aeroflow Health, explains, “Sizes would come in in different data formats, and we couldn’t just pass it to our internal orders table and submit the order automatically. It would land on one of us internally to process those orders manually. But the AI agent allowed us to decipher all this data into one standardized format that we could automate. That means we can complete the order in a two-way text conversation.”

Aeroflow’s Mom & Baby customers are new and expectant mothers—to put it mildly, they are not always awash in free time, nor time to sit down, nor even full of moments with two free hands. Being able to repurchase in a few seconds, with a few taps in a text thread, radically reduces friction and gets them the medical devices they need.

The results

For Aeroflow, the results have been remarkable—a 8.8% lift in repurchase conversions over business as usual. Brian has plans to deploy the BrazeAI Agent Console™ in a variety of other use cases, embedding flexible conversational agents across the customer lifecycle.

As Aeroflow’s results show, brands are already using agentic workflows to deliver value to their customers. This value comes from using agents which are embedded as steps in their Braze Canvas. The agents are units of composable intelligence that allow the Canvas to take in unstructured data like natural language conversation with customers, and respond fluently, in real time, and at scale. For Aeroflow, the returns on their innovation and investment in agentic marketing are remarkable, but they’re only the beginning.

Learn more about BrazeAI™ today.