My AI Lazzie encounter: How Lazada’s AI makes shopping smarter, faster and more human

THERE was a time when online shopping felt like roulette. You clicked through a sea of listings, hoped the seller was legit, prayed the courier arrived on time and crossed your fingers that the return policy worked in your favor. For tech buyers — such as myself, who routinely tested gadgets and accessories — those leaps of faith were occupational hazards. As a tech buyer, you bought, you waited and you hoped.

Then, I met AI Lazzie, Lazada’s artificial intelligence (AI) shopping assistant. What I expected to be another templated chatbot turned out to be something different — intuitive, conversational and quietly observant. It didn’t just reply; it anticipated. AI Lazzie remembered my browsing patterns, noted my wish list and refined its recommendations each time I returned.

When I searched for a new pair of studio headphones, Lazzie didn’t drown me in product listings. Instead, it compared features, summarized user reviews and even explained sound differences — from bass response to detachable cables — all in conversational shorthand. What used to be an impersonal scroll suddenly felt like a guided conversation.

This wasn’t browsing. This was curation — data-driven, context aware and human in tone.

Shopping without friction

Lazada’s AI Lazzie was part of a growing ecosystem of intelligent agents that shaped the entire e-commerce journey. Powered by Alibaba Group’s Qwen large language model, AI Lazzie has evolved into what Lazada called a “conversational co-pilot”. The system didn’t just process queries; it understood intent and context. In its pilot phase, Lazzie reportedly increased orders by 42 percent and drove 50-percent more user interactions.

PERSONAL SHOPPER AI Lazzie, Lazada’s new artificial intelligence (AI) shopping assistant that reads consumers’ browsing behavior, tracks their wish lists and offers recommendations that feel surprisingly intuitive as attested by The Manila Times tech editor Jing Garcia. IMAGE FROM LAZADA

This 11.11, AI Lazzie introduced SmartStack, a feature that let shoppers combine LazRewards and seller vouchers in one go. There was no more code-hunting required; you just typed “SMARTSTACK” in the chat, and the system did the math.

But AI Lazzie was only the face of a broader transformation. Beneath the interface, Lazada has deployed four other agents — a Refund Agent, Logistics Agent, Product Listing Agent and Marketing Agent — each working autonomously but collaboratively to remove friction at every step of the shopping experience.

Howard Wang, Lazada Group chief technology officer, called it the dawn of “agentic commerce,” a model where intelligent agents understood, analyzed and acted in real time.

“We’re moving from using AI as a support tool to empowering it as a co-pilot,” Wang said. “By integrating intelligent agents across our ecosystem, we’re making this 11.11 not just bigger but smarter — where shoppers enjoy greater convenience and sellers benefit from sharper insights and efficiency.”

That “smarter” manifested in real ways.

When one of my packages arrived with a damaged cable, I decided to test AI Lazzie’s problem-solving skills. The return process — once a waiting game — was completed in minutes. The Refund Agent verified my claim, generated the return code and scheduled the pickup automatically. The refund appeared in my Lazada Wallet the next day.

In the background, the Logistics Agent tracked the replacement item, processed inquiries in as fast as 20 seconds and sent real-time updates. Lazada said 35 percent of refund requests were now handled autonomously with a 99-percent accuracy rate.

This wasn’t customer service in the old sense; it was AI-assisted empathy — always on, never impatient and unbothered by queues.

When the buyer becomes the seller

Curiosity led me to try Lazada’s AI tools from the other side of the marketplace — as a seller. Anyone who has managed an online store knew the grind: writing product descriptions, editing images, tracking prices, and adjusting inventory. Lazada’s new Product Listing Agent changed that equation completely.

The AI auto-generated titles, translated descriptions across multiple Southeast Asian languages and even suggested competitive price ranges based on real-time demand. In just a few clicks, my listings were optimized for visibility with no need to guess the right keywords.

The impact was measurable. Lazada reported that the Listing Agent has already optimized more than 13 million listings, saving sellers an average of 11 work hours and increasing page views by up to 180 percent within a week.

Complementing that was the Marketing Agent, aptly named Sponsored Max, which used real-time algorithms to match ads with shopper intent. For small merchants, this was transformative — machine intelligence doing what marketing departments used to handle manually.

In Southeast Asia, where e-commerce competition was fierce, such tools were democratizing access to enterprise-grade analytics. Lazada data showed that 89 percent of sellers saw AI as a productivity booster, and 93 percent viewed it as a cost-saving enabler, but only 37 percent had actually adopted it. Bridging that gap, Lazada said, was central to its AI playbook.

The “human” algorithm

For all its automation, what stood out about AI Lazzie was its restraint. It didn’t overwhelm; it assisted. It listened. The tone was conversational, the timing was precise, and the language was warm. AI Lazzie didn’t try to sell; it tried to help.

Behind the interface, Alibaba Cloud’s AI infrastructure orchestrated everything from supply chain predictions and delivery route optimization to warehouse allocation and demand forecasting. The result was a platform that blended machine precision with something increasingly rare in e-commerce: empathy.

Ironically, the most “artificial” thing about Lazada’s AI ecosystem was also its most human trait: the ability to adapt, to understand nuance and to learn from experience.

That was the paradox of this new retail intelligence. The real innovation wasn’t mechanical; It was emotional.

Lazada’s five AI agents — spanning discovery, refund, delivery, product listing and marketing — were less about replacing people and more about amplifying them. They handled the repetitive tasks in order that humans could focus on the relational.

The new normal of e-commerce

After weeks of using AI Lazzie, I realized I wasn’t just shopping differently; I was thinking differently. Decisions came faster. Refunds were seamless. Recommendations felt genuine. The platform no longer felt like an app but a partner; one that learned, adapted and evolved with me.

This was part of Lazada’s commitment to making the platform the premier destination for authentic brands and high-quality products, giving shoppers access to exclusive global assortment and making shopping smart with AI.

This was what the future of e-commerce looked like when AI became invisible, when technology stopped shouting and started listening.

In an age where digital trust remained fragile, AI Lazzie and its agentic counterparts showed how artificial intelligence could make commerce not just smarter but more human.

For all its code and computation, Lazada’s AI proved that empathy could, in fact, be engineered.

TECH EDITOR Orlando 'Jing' Garcia II is the Sunday Business & Information Technology section editor of The Sunday Times for The Manila Times . He writes on digital transformation, artificial intelligence and Southeast Asia's emerging business tech economy. IMAGE FROM JING GARCIA