When we walked into the Cambridge Judge Business School Vibe Coding Hackathon on a Saturday morning, all we really had was a shared frustration and a blank page. By the end of the day, we had Commenti: an all-in-on review intelligence platform for e‑commerce sellers, built in under eight hours, and selected as the winner of the Hackathon.
We are a team of five students from Southeast Asia, a region full of fast‑growing digital businesses and equally fast‑growing problems. E‑commerce has exploded over the past decade on platforms such as Shopee, Lazada and TikTok Shop. Many of these sellers are solo founders or very small teams, running everything from product sourcing and marketing to logistics and customer service.
That is where we saw the gap: reviews. Reviews function as digital word of mouth, shaping trust and conversion online, yet the entrepreneurs we know rarely have the time to monitor every product page, across every platform, and then respond thoughtfully. Negative comments go unanswered, valuable suggestions are missed, and reputations quietly erode. Drawing on his experience as a product marketer in fintech, Chayanan channelled a familiar frustration: ”I had seen so many good customer insights get lost in feedback threads. With Commenti, I wanted those insights to finally drive better products”.
Our team brings the best of both worlds, with engineers who can actually code and business builders who have taken products to market before, giving us the right balance to design with customer, commercial and technical perspectives in mind.
Zern Nattapat, who has previously built ventures in Metaverse and Web3, kept reminding us “Every feature we designed started with one question. Would a real seller in Southeast Asia actually use this, and would they pay for it?”
Commenti became our response to that shared pain point. Over the course of one intense day, we designed a platform to centralise reviews across TikTok Shop, Shopee, Lazada and more, to flag urgent negative comments, and to generate on brand AI replies for neutral and positive feedback. We also explored how to detect spam or abusive content so sellers could focus on genuine customers. Our vision was simple: save time, protect brands, and turn customer reviews into growth. For Timotius, who has done QA before, the challenge was to make sure this vision actually worked under pressure: ”Even in a hackathon, reliability mattered. If a seller cannot trust the alerts or the responses, the product fails, no matter how clever the idea is”.
The most transformative part of the experience was seeing how accessible product building has become. In the past, a project like this would have required a fully technical team and weeks of engineering effort. With GenAI tools and vibe coding, we moved from idea to prototype in hours. Product creation is no longer reserved for those who can write complex Python or JavaScript from scratch. As Natasya, who had never coded before, reflected, ”I could focus on crafting a smooth, intuitive experience, and AI helped us translate that into a real interface. It felt like the product was finally being built around the user, not around the code.”
Reflecting on the hackathon, we left with more than a prize. We gained a lived understanding of how AI can democratise product building, how cross functional teams can compress months of work into a single day, and how real world pain points from our home region can inspire globally relevant solutions. Mew Phattanun, our main coder, emphasises the importance of solid logic and sees AI as a powerful tool rather than the brain behind the product, ”AI did not replace the need for logic, it amplified it. Our ideas still had to be technically coherent and actually work.”
After the Hackathon, this was proof that if you deeply understand a problem, you can now help build the solution, whether or not you call yourself a ‘tech person’.
It feels like the beginning, not the end, for both Commenti and for us as builders.