I have always felt a personal connection to Vietnam. Part of it comes from the familiarity of seeing Chinese characters in certain places, but most of it comes from the country itself: the food, the culture, and the warmth of its people. Over the past two years, as I have visited the country multiple times for developer workshops, hackathons, and community programs, my strongest impression has come from its developer community.
Vietnam has long been known internationally as a strong software outsourcing market. That reputation is still valid. The country has built a deep base of engineering talent by supporting global technology companies and software projects. But what I have seen across repeated visits is a shift in mindset. More young developers are no longer thinking only about working for global companies or delivering software for others. They are starting to think like builders.
My first major developer trip to Vietnam was in June 2024, when I attended Tech in Asia Saigon, where Agora was a Platinum Partner. At that time, AI was already beginning to emerge as a major global topic, but most local developers I met were still focused on more familiar technologies such as real-time communication, or RTC, and traditional software stacks.
AI was widely discussed, but it still felt distant from day-to-day production work. Many developers saw it as important, but not yet practical enough to reshape their immediate careers or workflows. Some believed the idea that AI could significantly change software engineering, or affect the role of junior developers, was still years away.
I remember speaking with one developer who asked whether it was still worth becoming a junior software engineer. It was a difficult question. At the time, my honest view was that it might not be the easiest moment to make that transition. He chose to pursue that path anyway.
When I returned to Vietnam in March 2026 for Lotus Hacks, the landscape felt very different. Nearly every developer I met was using tools such as Cursor, Claude, Codex, or similar AI coding assistants. AI was no longer something being evaluated from a distance. It had become part of the default development workflow.
The most important shift was not that Vietnamese developers had discovered AI. It was that the debate had already moved on. They were no longer asking whether they should use AI. They were asking how to use it better, and personally, I think that distinction matters. It suggests a move from adoption to fluency.
When AI stops being the headline
One of the clearest signs of this shift at Lotus Hacks was that AI itself was no longer the differentiator.
Many teams were using similar AI infrastructure layers and tools. In an earlier cycle, simply adding AI to a product might have felt novel. By 2026, that was no longer enough. The stronger teams were not defined by whether they used AI, but by whether they understood the problem they were trying to solve.
That is a healthy sign for any ecosystem. It means the conversation is moving from technology-first thinking to problem-first thinking.
Lotus Hacks itself reflected this broader momentum. The event attracted more than 1,000 developers, more than 200 project submissions, and participants from more than 30 countries. What surprised me most was not only the size of the event, but its international pull. I met developers from Malaysia, Singapore, Canada, Finland, and several other countries who had travelled to Vietnam to participate.
That changes how we should think about Vietnam. It is not only a place where global companies source engineering talent. It is increasingly becoming a place where builders want to gather.
From wrappers to real products
The evolution of project quality has also been clear.
In 2024, I did not see many AI-native projects in Vietnam. AI still felt far away from most developers. By 2025, during a conversational AI workshop, more GPT-based applications began to appear, often as wrappers with an added user interface. That was a natural first step. Developers were learning what the tools could do.
By 2026, however, the projects had become more grounded. More teams were building around industry needs rather than abstract AI capabilities. In hackathon submissions and community discussions, I saw projects tied to manufacturing, logistics, e-commerce, livestream commerce, education, and agriculture.
This is where Vietnam’s AI ecosystem becomes particularly interesting. Its startup ideas are not simply mirroring Silicon Valley trends. They are being shaped by the country’s own industrial structure.
Vietnam has a strong manufacturing base, a fast-growing digital economy, active e-commerce adoption, and large practical needs across education, logistics, and agriculture. Google, Temasek, and Bain & Company estimated Vietnam’s digital economy at US$39 billion in 2025. That growth gives local developers more real problems to build around.
Not every AI product needs to be global from day one. Some of the most meaningful products may be narrow, local, and deeply practical.
A country with two realities
One of the most striking things about Vietnam is the coexistence of two very different realities.
On one hand, much of the economy is still rooted in traditional manufacturing, offline commerce, agriculture, and service work. Many people have limited direct exposure to AI in their daily lives. On the other hand, a fast-moving group of young developers now sees AI as a direct path to mobility, independence, and entrepreneurship.
Two years ago, many developers I met were focused mainly on stable employment, whether in outsourcing firms or larger technology companies. That remains true for many people. But compared with my earlier visits, I now see a larger group actively exploring side projects, startup ideas, and AI-native products.
AI has lowered the cost of experimentation. A small team, or even a single developer, can now prototype something that previously required a much larger engineering group. For young Vietnamese developers, this makes entrepreneurship feel less distant. AI is not only seen as a threat to jobs. In many conversations I had, it was seen as a tool to accelerate careers, test ideas, and possibly change one’s life trajectory.
This does not mean the transition will be smooth. AI can create uneven outcomes. It can widen the gap between those who know how to use the tools and those who do not. But in Vietnam, I have seen a strong sense of agency among young builders. They are not waiting for perfect conditions.
Ho Chi Minh City’s builder density
Vietnam’s developer ecosystem is still highly concentrated in major cities. I plan to spend more time in Da Nang and Hanoi this year, and I already have developer friends in both places. But from my visits so far, Ho Chi Minh City clearly acts as the dominant gravity center for talent, startups, consumption, and community activity.
This concentration creates both advantages and risks. A dense ecosystem accelerates learning. Developers meet each other more often, events are easier to organize, and ideas move faster. Startups, investors, mentors, and communities become easier to connect.
At the same time, a strong center can pull attention and resources away from other regions. If Vietnam’s AI ecosystem is to become more inclusive, other cities will also need stronger community infrastructure.
Still, Ho Chi Minh City’s momentum is hard to ignore. StartupBlink’s 2025 ranking placed the city among Southeast Asia’s top five innovative startup ecosystems. That matches what many people in the community already feel on the ground: density is turning into momentum.
Policy gives direction, but communities create speed
Vietnam has already set a national AI ambition. Its national AI strategy, issued in 2021, aims to place the country among the top four ASEAN countries in AI research, development, and application by 2030. That direction matters because it gives the ecosystem a clear signal.
But from what I have seen, the fastest movement is coming from the ground up.
Young developers are experimenting faster than institutions can adapt. Communities are forming around tools, projects, and events. Local developer leaders, cloud ambassadors, startup founders, and AI enthusiasts are helping translate global AI momentum into practical learning for Vietnamese developers.
This grassroots acceleration is closely tied to Vietnam’s outsourcing legacy. For years, the country built a strong base of engineering talent by serving global software needs. That history should not be dismissed. In fact, it may now become the foundation for the next phase.
The same engineering discipline that supported outsourcing can now support product building. The difference is mindset. Developers are beginning to move from delivery to ownership.
Another signal worth watching is the growing attention Vietnam is receiving from global technology companies. Chinese and American technology companies may compete intensely in many markets, but many appear to be reaching a similar conclusion about Vietnam: it has the talent, energy, and practical market needs to become one of Southeast Asia’s most promising AI builder ecosystems.
When ecosystem builders from different regions arrive at the same conclusion independently, it is usually because they see the same underlying opportunity.
The interface matters as much as the intelligence
Vietnam’s AI opportunity will not be defined only by models, APIs, or developer tools. It will also depend on how people interact with AI.
Take agriculture as an example. AI can help detect crop disease from images, improve access to information, or support decision-making for farmers. But many end users may not be comfortable typing prompts or navigating complex interfaces. They may prefer speaking. They may use regional accents. They may switch between languages or dialects.
This is where voice becomes important. In developer conversations around Voice AI, one recurring point is that language and accent support are not minor technical details. In Vietnam, differences between Northern and Southern Vietnamese accents can affect whether an AI product feels natural, usable, and accessible.
The interface matters as much as the intelligence. If AI is going to move beyond developers and early adopters, it must fit the way people already communicate.
What Vietnam could become next
Vietnam is not yet like the US, China, Japan, or South Korea, where advanced technology is deeply embedded in many layers of daily life. The connection between AI and everyday consumer experience is still developing. That is precisely why the opportunity is interesting.
Vietnam’s next wave of AI startups may not all become global giants. But many could solve meaningful local problems: helping small businesses operate more efficiently, supporting farmers, improving education access, making logistics smarter, or creating more natural digital services for Vietnamese consumers.
When I return to Vietnam for another developer event this July, I will be paying less attention to the size of the gathering and more attention to what people choose to build. That is where the real signal is.
From my perspective as someone who visits periodically rather than lives there continuously, the shift in Vietnam does not feel incremental. It feels structural.
Vietnam is no longer only an outsourcing hub. It is becoming a country where developers are starting to think like builders. And if that mindset continues to spread, Vietnam’s most important technology story may not be the talent it provides to the world, but the products, companies, and ideas it builds for itself.
Derek Zheng is APAC Lead of Developer Relations and Director of Customer Success at Agora.
TNGlobal INSIDER publishes contributions relevant to entrepreneurship and innovation. You may submit your own original or published contributions subject to editorial discretion.
Featured image: Conny Schneider on Unsplash