Talk Commerce Talk Commerce
Back to Blog
Luca Borreani
Podcasts
| 7 min read

AI Agents Are Transforming E-commerce Customer Support and Operations

By Brent W. Peterson

In this episode of Talk Commerce, host Brent Peterson sits down with Luca Borreani, co-founder and CMO of ZipChat AI. The conversation explores how AI agents are reshaping the e-commerce landscape, moving beyond simple chatbots to become autonomous entities capable of handling complex customer interactions and business operations. Luca shares his journey from affiliate marketing to building enterprise solutions and offers practical insights for merchants looking to implement AI in their businesses.

Key Takeaways

  • AI has evolved from simple generative chatbots to autonomous agents that can take action on behalf of merchants
  • Everything in modern AI platforms should be promptable, allowing merchants to use natural language to set parameters and guardrails
  • Merchants are now openly disclosing when customers are interacting with AI agents, marking a shift from earlier attempts to disguise automation
  • AI agents can handle operational tasks like order confirmation, upselling, and even initial refund assessments
  • The technology compounds over time, learning more about customers, operations, and brand identity with continued use
  • Black Friday and Cyber Monday present opportunities for AI to analyze customer conversations and provide insights for better targeting

About Luca Borreani

Luca is a serial entrepreneur who began his journey in affiliate marketing while still in university. After achieving early success promoting other companies’ products, he and his partner launched their own e-commerce ventures, initially in dropshipping. Recognizing the pain points in that industry, Luca co-founded UDROPPY, which received funding from Sequoia Capital and Jason Calacanis after initially struggling to attract Silicon Valley investors.

Following the sale of his shares in UDROPPY, Luca joined forces with a former customer who was tackling AI-powered customer support challenges. Currently based in Dubai, he serves as CMO of ZipChat AI, where he benefits from AI tools that make marketing operations more efficient than in his previous ventures. His background in e-commerce gives him unique insight into the problems merchants face and how AI can solve them. Luca Borreani remains committed to making AI accessible and practical for businesses of all sizes.

Episode Summary

The conversation begins with Luca explaining his unconventional path to Dubai. After completing his second master’s degree, he and his business partner were running profitable affiliate marketing campaigns. They wanted to move to New York but faced visa complications. Dubai offered proximity to Italy, straightforward bureaucracy, and excellent connectivity between Europe and Asia.

Luca traces his entrepreneurial arc from affiliate marketing to dropshipping, then to founding UDROPPY. “If you are good at promoting other people’s stuff, why don’t we promote our own stuff?” he recalls thinking. The dropshipping business revealed systemic problems with payment processors and shipping times from AliExpress, which inspired UDROPPY’s creation.

When discussing AI agents, Luca draws a clear distinction between early generative AI chatbots and current autonomous systems. “AI is much more than that. It’s not just a reply. It’s potentially an autonomous entity that can take action on your behalf or on behalf of the merchant,” he explains. Unlike Shopify’s integration with OpenAI that allows customers to purchase through ChatGPT, ZipChat focuses on merchant-side automation.

The platform allows merchants to automate repetitive tasks in customer support and success. These include following up on questions, recovering abandoned carts, editing orders, issuing refunds, and making product recommendations with upselling and bundling opportunities. Luca emphasizes that merchants maintain control through natural language prompts that define when, where, and how the AI should act.

Addressing concerns about AI reliability, Luca notes that merchants set their own guardrails. “You literally use normal language to describe what you want it to happen, and the AI will interpret it and act on it,” he says. For example, merchants can specify discount limits for negotiations, ensuring the AI won’t offer excessive discounts while trying to close sales.

The conversation shifts to transparency in AI interactions. Luca observes a significant trend change over the past two years. “When we started two years ago, everybody was trying to trick the final user,” he admits. “They were trying to never say it’s an AI, because they felt like people would not like to be talking to AI.” Now merchants typically disclose they’re using AI assistants, often with a simple disclaimer stating the AI represents the brand.

ZipChat handles escalations by transferring conversations humans can’t resolve to ticketing systems or allowing human takeover within the platform. The system includes anti-spam filters and automatically recognizes when conversations are unproductive. “If the AI sees the conversation is going nowhere, it basically starts to cut off,” Luca explains.

For operational tasks, ZipChat takes a cautious approach. The platform excels at confirming cash-on-delivery orders, which are prevalent in Europe, Asia, and Africa. Instead of expensive call center operations, the system sends automated WhatsApp messages that feel personal and follows up based on order details and conversation history. “The message will be really one-on-one based,” Luca notes. The system can also update orders and add upsells based on customer responses.

When it comes to more sensitive operations like refunds, Luca advocates for maintaining human oversight. “There are still a few things that even if it’s a mistake only one time, still it’s something you don’t want to happen,” he acknowledges. The platform is testing features where customers can send photos of damaged products, but final refund decisions still require human approval.

Looking ahead to Black Friday and Cyber Monday, Luca highlights AI’s ability to analyze months of customer conversations to identify trends. Merchants can understand what offers customers expect, what upsells make sense, and which marketing angles will resonate. “They know already what kind of offers or upsells people are expecting based on the past six, nine months of conversations,” he explains.

During high-volume periods, AI agents provide immediate value by proactively engaging customers, understanding their needs, and answering pre-purchase questions. “Most customers just want to be heard. They want to feel like on the other side, there is an interaction and a real entity or something real happening,” Luca observes. This immediate engagement helps convert traffic that might otherwise be lost during competitive shopping periods.

Luca concludes by encouraging merchants to experiment with AI despite any intimidation they might feel from complex workflows they see on social media. “Somebody already figured it out how to apply to your business in an easy and kind of plug and play way,” he reassures. The technology’s compounding nature means early adoption pays dividends. “The longer you’re going to be using it, the better it’s going to be.”

Final Thoughts

AI agents represent more than just another marketing tool or customer support solution. They’re becoming integral to e-commerce operations, handling everything from customer conversations to order management. The technology’s ability to learn and improve over time makes it increasingly valuable for merchants willing to implement it now. As Luca Borreani demonstrates, the key isn’t creating the perfect AI system from day one but rather starting the journey and allowing the technology to compound its understanding of your business, customers, and operations.

Are you ready to let AI agents handle your customer conversations, or will you wait until your competitors gain the conversational advantage?

Follow Talk Commerce on your favorite platform:

Tags

AI agents AI transparency autonomous AI Black Friday strategies chatbot technology conversion optimization customer service automation customer support AI e-commerce automation e-commerce operations Luca Borreani merchant tools order management AI ZipChat AI

Author

Brent W. Peterson

Brent W. Peterson

View all posts by Brent W. Peterson
Back to all posts