Skip to Main Content

Jan 30, 2025 | 3 minute read

Adding AI Agents into Your Commerce Experience? Make it Natural

written by Bryan House

Retailers are experimenting with all kinds of AI applications for their online storefronts. But, which ones matter most to actual shoppers? Recent research from Bain & Company shows that 71% of shoppers were unaware they were interacting with AI, even though they had recently visited retailers using generative AI technology. When asked about barriers to adoption, most customers said that their current, non-AI shopping methods were just fine, or they had a perceived lack of need for AI.

Now, if you flip the script and ask a marketer or merchandiser, AI’s potential is transformative. The market share for AI in e-commerce is expected to expand at a nearly 25% compound annual growth rate (CAGR) in the next 10 years. From optimizing product detail pages with generative AI to launching dedicated AI agents as personal shoppers, marketers are rightfully casting a wide net when it comes to investing in AI. I’m always of the mind that if a technology differentiates your brand, then it’s worth the investment.

But, with the fast-growing breadth and depth of AI capabilities, where do you start? My best advice is to make it helpful to the buyer, but most of all, make it natural for them to use.

Experimenting with AI Agents on Elastic Path

The concept of agentic AI is reaching a peak in the hype cycle. We’ve been slowly moving in the direction of having AI agents answer our questions and even act independently on our behalf. The impact of agents on the shopping experience could be massive, especially for companies with a lot of customization options and variations of their product.

Elastic Path is no exception here. As a composable commerce company, there are a lot of different permutations of our product, along with highly customizable deployment patterns that depend on the customer’s problems to be solved. With so many options to start your composable commerce site, we wanted to create an AI agent to help customers and prospects navigate through our product documentation.

Our goal with Elastic Path AI is to make it much simpler to get answers to questions about our product. All too often, product enablement follows a checklist: “Take this course, then take a test, and you’ll get a certification badge for LinkedIn.” Often, a general course doesn’t help actual users achieve what they’re trying to do in the product. With Elastic Path AI, we want to help our users effectively navigate the large corpus of knowledge in our documentation and enablement, getting straight to the answer for the question they are asking right now. Whether that’s code samples or instructional content on how to take a certain action in our product, the agent can provide users with very specific next steps.


In the future, we may take this agent a step further and have it independently take action based on the user’s request and product experience. There are certainly a lot of possibilities we can look at, but we want to first test and learn from the initial deployment of Elastic Path AI, making it as natural as possible for our customers to ask questions about our documentation.

The good news? Composable commerce architectures, especially those like ours that adhere to the OpenAPI spec, make it much easier to try new technology and create differentiating experiences for customers — often with low risk and high reward. Instead of spending months (or even years) on custom development work, APIs lower the barrier to entry for experimentation, testing, and learning with all kinds of AI technology, including generative AI.

Here’s to creating a culture of commerce experimentation in 2025!

Stay up-to-date with Elastic Path

Sign up to hear more about commerce, merchandising and development best practices, and our flexible, API-first commerce platform.

Loading Form...