Retail is stepping into a new era, one where conversation replaces clicks and shoppers expect brands to listen, rather than sell. This shift toward conversational commerce is rewriting the rules of engagement. Instead of static product pages and one-way campaigns, consumers now crave real-time, human-like interactions that feel as intuitive as chatting with a store associate.
At ContactPigeon, we saw a unique opportunity. What if retailers could merge their powerful customer data with a conversational agent that can hold an actual conversation like a seasoned sales rep? That vision became Menura AI, a retail AI assistant designed to blend intelligence, empathy, and persuasion into every interaction.
Menura is the voice and brain of a new shopping experience, one that learns to assist, personalize, and convert like a human. In this behind-the-scenes story, we’ll share how Menura was imagined, built, and refined, from a spark of inspiration to a full-fledged AI for retail innovation.
Why retail needed a smarter AI assistant
Every great idea starts with a frustration, and in retail, there were plenty. Data was fragmented, scattered across channels. Campaigns were one-size-fits-all, pushing messages instead of creating experiences. Checkout drop-offs remained stubbornly high, and customer journeys felt anything but personal. Meanwhile, chatbots were answering FAQs without understanding the products, the customers, or the nuance behind a shopper’s intent.
At ContactPigeon, we had already tackled many of these challenges with our Customer Data Platform (CDP), helping retailers unify data, analyze them and act at scale with personalized automation. But even that wasn’t enough, because retail itself was changing. Shoppers no longer wanted to be “marketed to”, they wanted conversations. They expected empathy, instant assistance, and contextual understanding that felt as natural as talking to a real store associate. No existing AI for retail solution truly knew what it was selling, beyond text descriptions and keywords.
We’d seen this gap firsthand. Years of working alongside retailers taught us one powerful truth: context, not campaigns, drives loyalty. Real engagement happens when every interaction feels relevant in the moment, not pre-scheduled or generic.
That realization sparked the idea for Menura AI. What if a retailer’s best in-store associate, the one who remembers every detail, senses hesitation, and knows when to upsell, could exist online, 24/7? What if that assistant could be trained on real product insights, customer behavior, and omnichannel interactions? From that spark, Menura AI began to take shape, an intelligent, conversational engine designed to transform retail AI personalization and make conversational commerce truly human.
Training Menura AI to understand products and speak retail
The challenge
If there’s one thing we learned early on, it’s that retail data is rarely clean and tidy. Product catalogs are messy and inconsistent, categories vary from brand to brand, and descriptions often range from overly detailed to frustratingly vague. One store might list heel height, materials, and color variants, another might simply say “elegant women’s shoes.” A PC might include full specs, while another only says “fast and powerful.” For any AI for retail, this inconsistency is like trying to learn a new language with half the words missing.
We’d already seen how these limitations affected personalization. Thus, to make Menura AI coherently understand products, not just read about them, we had to go beyond the text. We needed to train Menura so that it could read between the lines, infer, and contextualize every item the way a skilled in-store associate does.
The approach
Our engineers designed Menura to not only ingest existing catalog data, but to expand and enrich it autonomously. The foundation starts with the XML feed each retailer provides, including core details such as product name, category, price, image, and URL. From there, Menura crawls the retailer’s website, extracting every available product detail: descriptions, stock availability, variations, and anything else that can help it form a full picture.
When text data isn’t enough, Menura turns to images. It analyzes visuals to infer attributes like size, shape, material, or even the probable product category, enriching the catalog with AI-generated tags and categories for better classification and search accuracy. Over time, Menura builds its own contextual understanding of each category, identifying shared traits (like heel height for shoes or CPU type for laptops) that define shopper intent.
To keep catalogs fresh, Menura automatically updates every six hours, checking for new products, price changes, and stock updates. Retailers can also trigger a manual refresh whenever new data is added. And for specialized sectors, say, food retailers with allergen details, Menura can even extract information from ingredient images.
This continuous learning cycle is supported by ContactPigeon’s omnichannel customer engagement platform, combining structured product data, conversational insights, and anonymized shopper interactions. A hybrid model, part natural language understanding (NLU), part behavioral prediction, helps Menura connect how products are described with how customers actually shop.
Finally, human-in-the-loop training ensures that Menura’s tone, timing, and recommendations align with retail best practices. Each iteration brings it closer to speaking the universal language of retail, empathetic, adaptive, and intelligent.
Giving Menura AI your brand’s voice
Before Menura AI could talk to shoppers, it needed to find its own voice. The name Menura comes from the lyrebird, nature’s ultimate mimic. Known for its ability to imitate any sound it hears, the lyrebird became our perfect symbol for adaptability and learning. Just like its namesake, Menura AI was designed to mirror each brand’s unique tone, adapting seamlessly from luxury elegance to playful friendliness.
But Menura doesn’t just “copy”, it learns. Retailers can shape its behavioral foundation by defining when and how it engages. Through custom prompts and triggers, brands can tell Menura when to offer a discount, when to suggest an accessory, or when to simply say, “Let me check that for you.”
More importantly, brands can inject a personality directly into Menura. A premium fashion brand might have it greet shoppers with refined warmth (“Welcome back, shall I suggest what’s trending this season?”), while a playful toy store might prefer a cheerful tone (“Ready to find something awesome today?”). In essence, Menura’s voice isn’t ours, it’s yours, fine-tuned for your audience and your story.
Learning in the wild
Once Menura AI could think and speak retail, it was time to see how it performed in the wild. We piloted the system with select ContactPigeon retail partners, starting with leading brands in consumer tech. The results spoke for themselves, higher engagement rates, a noticeable recovery in abandoned carts, and an uplift in average order value (AOV).
Each real shopper interaction became a lesson. Through a constant feedback loop, Menura refined its tone, adjusted timing, and learned when to listen rather than talk. As we expanded to beta users, a new learning phase began, one focused on tone perception, latency, and accuracy, all fine-tuned in weekly model cycles. In the end, the field tests confirmed what we always believed: the best retail AI learns empathy from feedback, not just data.
What’s next for Menura
Menura’s story is just getting started. What began as a conversational assistant for eCommerce is now evolving into a complete retail engagement companion, one that meets customers wherever they are. Our roadmap extends far beyond on-site chat. Soon, Menura AI will seamlessly connect through third-party messaging apps like Viber and WhatsApp, bridging the gap between in-store service and digital convenience.
The next frontier? A more agentic AI, one capable of making autonomous decisions while staying true to each brand’s tone and customer intent. By leveraging ContactPigeon’s deep behavioral insights, Menura will anticipate needs, take proactive actions, and adapt in real time without losing its human touch.
Our long-term vision is bold yet simple: to build a retail ecosystem where conversation, prediction, and personalization work as one. In that future, Menura is not an assistant anymore, but the heartbeat of intelligent, empathetic retail engagement.
