Would your organization build your own digital infrastructure – including AI governance and cybersecurity – all on your own?

Coolmate, a rapidly growing Vietnamese direct-to-consumer (D2C) apparel brand, has quietly built much of its digital infrastructure inhouse as it scaled across multiple e-commerce marketplaces and, more recently, into offline retail.

Rather than relying entirely on off-the-shelf platforms, the company developed its own e-commerce stack and internal ERP to manage inventory, sales and operational data across channels. We find out why and how from Hiep Nguyen, co-founder and CTO, Coolmate.

Why did Coolmate choose to develop its own e-commerce stack and internal ERP to manage inventory, sales and operational data across channels, rather than deploying an off-the-shelf solution?

Nguyen: At Coolmate, we see technology and data not simply as support functions, but as a core part of how we build efficiency, trust and customer experience at scale.

From the outset, our business model has been direct-to-consumer. As we began scaling across multiple digital channels, we quickly realized that most off-the-shelf e-commerce solutions were not designed to support the operational complexity of managing multiple marketplaces while maintaining tight control over inventory, costs and customer experience.

Building our own e-commerce platform and internal ERP allowed us to take full ownership of our operational backbone. Today, this centralized system manages inventory, sales and customer data across our website and marketplaces such as Tiki, Lazada, Shopee and TikTok. It enables us to synchronize stock in real time, prevent overselling or overstocking, and maintain a consistent customer experience across channels.

Developing our own stack has also given us the flexibility to scale and adapt quickly. When we expanded into physical retail in 2026, for example, we were able to integrate our new stores into the existing system within just a month, without disrupting operations.

Ultimately, owning our technology infrastructure gives us the agility to innovate while ensuring that the data and insights underpinning our business remain secure, structured and actionable as we continue to grow.

What are the challenges and advantages of building a proprietary in-house digital retail infrastructure?
Nguyen: Building a proprietary in house digital retail infrastructure comes with both challenges and advantages. On the challenge side, it requires significant investment in time and resources to design, develop and to maintain as the system handles everything from inventory and sales across multiple channels to customer management and reporting. It also demands technical expertise and careful planning to ensure reliability and security, particularly when handling sensitive data such as customer information and financial metrics.

The advantages, however, are substantial. Having our own system gives us full control and flexibility to tailor the platform to our exact business needs. It allows us to integrate online and offline operations seamlessly, manage inventory in real time, and respond quickly to changing market demands.

This centralized approach also supports data driven decision making, giving us insights that would be difficult to capture with off the shelf solutions. Ultimately, it enables us to scale efficiently while maintaining high standards of quality and service for our customers.

How do you handle AI and cybersecurity in the business, especially when managing sensitive customer and commercial data and applying AI across media production, software development and analytics?

Nguyen: At Coolmate, we take both AI and cybersecurity very seriously, particularly because we handle sensitive customer and commercial data.

On the cybersecurity side, our internal systems, including the CRM for our website and our ERP, run on physical on-premises servers rather than cloud services.

This allows us to control access tightly and manage software and protocols internally, while our external partner handles the physical security, firewalls and hardware. We also conduct yearly cybersecurity audits and red team testing to identify any vulnerabilities and strengthen our systems accordingly.

Customer data from marketplaces is not accessible to us, so we only manage data from our own channels. This data is kept in a separate database with access restricted to a very small number of system administrators. Sensitive information such as phone numbers or emails is hidden from staff, and all access is logged, so any abnormal activity can be identified and addressed.

Regarding AI, we apply it where it can add the most value. In media production, it helps in graphic design and video production; in software development, it allows our lean tech team to code and release features faster; and in analytics, it assists staff in scanning data and generating insights. AI has also been very useful in helping with product design.

Access to AI involving sensitive data is restricted to leaders who have been trained by our internal team, ensuring that the information is handled securely. Overall, by combining strong cybersecurity practices with careful and targeted use of AI, we can innovate, improve efficiency, and protect both our data and our customers.