'Startups should ask businesses what they need to start an AI business'

Experts advise AI startups to approach businesses to solve practical problems, both generating early revenue and understanding market needs.

Commenting at the discussion "Startups in the context of AI development" within the framework of "Tuoi Tre Startup Award 2025" on the afternoon of July 22, Mr. Le Hong Minh, Founder and Chairman of the Board of Directors of VNG said that AI "sounds very interesting, innovative and creative" but in reality "is very difficult to do", which is a big problem for startups.

He pointed out that the AI market changes very quickly, leading to the risk that startups will build products that are outdated by the time they launch. Therefore, we need another way to both make money right away and understand real needs.

"Instead of thinking and wasting time making a product that is not yet known to the market, you should start by approaching businesses directly and asking them: 'What problem are you facing right now? Hire me to solve it,' he recommended.

According to the head of VNG, many businesses are willing to spend money to solve specific problems with AI. Startups can accept projects at competitive prices but in return will accumulate knowledge, experience, and even intellectual property. "The opportunity is clearly there, the problem is your ability to convince them to trust you, and your ability to implement it in practice," said Mr. Minh.

Mr. Le Hong Minh, Founder and Chairman of the Board of Directors of VNG at the seminar. Photo by the organizers

Ms. Dang Huynh Uc My, Chairwoman of the Board of Directors of Thanh Thanh Cong Bien Hoa (TTC AgriS), said that the company has transformed from a sugar producer to a provider of comprehensive agricultural solutions using AI. According to her, businesses sometimes lack new perspectives and breakthrough energy to break out of the rut. "In that context, startups bring new thinking and flexible approaches," she said.

https://www.youtube.com/watch?v=Y03dtI1u8iI

The "28th Annual Global CEO Survey - Asia Pacific Region" - including Vietnam, conducted by auditing firm PwC, also showed that 82% of CEOs surveyed said they had put GenAI into operation in the past 12 months. In 2024, 37% said their revenue increased thanks to this technology.

Research and Markets (Ireland) forecasts that the size of Vietnam's AI market will increase from 2.15 billion USD (2023) to 23.06 billion USD in 2028, an annual growth of 48.5%. Statistics from PitchBook (USA) show that Vietnam had about 765 startups operating in the AI field last year, ranking second in Southeast Asia after Singapore.

For young AI startups struggling with the "bread and butter" problem, Mr. Dang Huu Son suggests using open source codes to quickly create MVP products in about 1-3 months, even using "no-code" tools (no programming required).

MVP (Minimum Viable Product) is to develop a product with just enough features to satisfy initial customers and collect feedback, as a basis for future product development.

"The important thing is to bring the product to market as soon as possible to have metrics, and then start looking for co-founders with expertise," said Mr. Son.

Mr. Nguyen Hoang Hiep, Director of Operations of AI Hay at the seminar. Photo of the organizing committee

In the capital raising stage, Mr. Nguyen Hoang Hiep, Operations Director of AI Hay - an AI-based knowledge search and research platform - recommends considering this activity as "going to sales". According to him, many startups are good at selling but think that raising capital is a completely different thing, so they ignore effective sales processes.

Specifically, for each type of investor such as curious, reserved or interested, startups must have their own strategy to convince. In other words, the experience of investors and customers is the most important factor that needs to be optimized.

"They want to see the big picture: What is your vision? Where is your advantage? Is the business model capable of generating revenue, even if there is no revenue at the moment? These are the questions that need to be answered. Let's save the technical details for later," he recalled his experience. Earlier this month, AI Hay completed a $10 million Series A funding round, bringing its total raised to $18.5 million.

Telecommunications