The next generation of e-commerce AI tools
When the marketing operations team at one of the world’s largest technology companies wanted to make it easier to navigate the complex portfolio of software subscriptions on their e-commerce website, they turned to iSoftStone.
The challenge: create an AI-powered product recommender for customers that combines the intuitive feel of conversational AI with the structured data and side-by-side visual ease of a comparison tool.
As Josh Welschmeyer, one of iSoftStone’s delivery directors, explains, “Our team already managed the client’s dynamic pricing interfaces, and we built their customer-facing AI assistant. Bringing those capabilities together for a product-chooser experience was the logical next step, and iSoftStone was a natural fit for the project.”
Together, using our client’s own suite of enterprise AI platforms, solutions, and software, we created an e-commerce AI tool that understands what a visitor is truly looking for and guides them towards the right product and pricing option for their needs.
An AI-powered product recommender at the cutting-edge
Like other large software companies, our client offers thousands of different products and subscription options across everything from cloud productivity suites to collaboration tools. In addition, each product typically has multiple tiers, licensing structures, and regional pricing models.
Customers reviewing static webpages must navigate details manually and jump around between product information pages. It’s a lot to wade through. Even with best-in-class non-AI comparison tools, the experience could still feel disparate and long-winded.
To create a compelling digital experience, any product recommender would need to parse all those granular details effortlessly and deliver smart conversational product discovery.
“AI is the right direction for showing options in a self-serve way,” says Welschmeyer. “By using an AI-powered product recommender, customers can ask questions and see what fits without digging through 50 different product pages.”
iSoftStone’s team integrated:
- Conversational AI platform and bot framework to manage logic and dialogue flows.
- Pre-trained AI models and enterprise AI platform to query the company’s extensive product catalog and interpret customer intent.
- Retrieval-augmented generation (RAG) training to improve the AI’s understanding of relationships between products, plans, and features.
At the front end, Adobe Experience Manager (AEM) handles the content and components needed for a sleek customer experience.
We see our team as a research and development arm for our clients. Everything we build for them shows what’s possible — and then we put it out into the world.
Josh Welschmeyer
Delivery Director | iSoftStone
An intuitive, self-serve experience powered by AI
The result is a hybrid experience — part conversational, part visual. Customers simply ask questions about what they’re interested in, and the interaction is effortless. They might prompt the AI with, “What’s the best productivity suite for a small business?” or “Which plan includes advanced security tools?”
The product recommender interprets each customer’s question and responds through a conversational AI assistant. Then, if needed, it guides the customer through a simple workflow of choices. The flow helps disambiguate products and define the customer’s needs. For example, they might select “personal” or “business” before the system presents a focused recommendation or a comparison interface. It’s quick, it’s seamless. And all within a clean, modern interface. The perfect AI-guided shopping assistant.
Different user preferences are accommodated, too. Some visitors may choose to rely solely on conversational AI to guide them; others may choose to use the comparison interface to filter options by feature, price, or plan type. Either way, it’s a streamlined, personalized journey, and it’s much easier for customers to make informed purchasing decisions based on clear comparisons.
And, for those who prefer, the option to browse manually, without AI, remains.
Smart conversational product discovery drives customer engagement
Our client is still tracking the full impact of the AI-powered product recommender, but so far, it’s delivering:
- Information hierarchy that makes it easier for customers to compare products and plans
- More transparency around the differences that can drive a purchase decision
- Reduction of repetitive content across static pages
- Increased buy and try engagement
- Growth in purchase completion rates.
While the AI-powered product recommender is in use only on one section of the client’s US website for now, they plan for it to become a keystone feature of sitewide UX in the longer term and be active across every market worldwide. We’re excited to see the impact it will make and, in the meantime, we’re already on to the next AI challenge for our client.
