Behind the claims about mastering artificial intelligence, many businesses and research unit in Vietnam still have to manage between ambition and reality, when the resource has not enough power to keep up the pace technology.
Between the office for a technology company in Cau Giay, Ha Noi, plates, plans, hanging sealed the piece of paper, multiple colors: each sheet is an idea for the project from chatbot customer care to platform data analysis automatically. But between the words excited her, there are sheet is crossed, below the caption curt: “delay – lack of personnel”.
Since the wave WHO wake late 2022, the company also welcomes the first movement, set a goal of “WHO” the whole products and development tool for end users. But when new ideas popping out every day, human resources become elements, drag them back to reality. “There are many projects, features forced to stop because there are not enough people deploy AI”, technical director of the company admitted.
Painting manpower in Vietnam WHO have many flourishes, but always in a shortage of both quality and quantity. In the annual report on artificial intelligence 2025 by the Institute of information technology – vietnam National University, Hanoi done months from January 7 to January 10/2025, the three groups surveyed, including business provides AI training facilities and units application are considered “human factor” is the biggest challenge for the development of artificial intelligence in Vietnam.
Out of 100 units surveyed, 45% of businesses provide WHO said the lack of high quality human resource is the biggest barrier, almost twice compared to the 23% of businesses facing difficult infrastructure compute and data.
“Teen who is good at making the job market WHO falls into a state of paradox: technology go ahead, people followed,” the leader of a research unit, WHO said.
Hunting talent artificial intelligence
“There are many ideas we want to do, but have to arrange the order of priority because of the lack of people,” Bui Thanh Minh, Deputy general director of Misa, said. Is one of the pioneer enterprise application WHO, at the same time get the task of building language models, big (LLM) and for those WHO “Make in Vietnam”, he said the company put out multiple objectives in which the transfer of the entire more than 50 products available to model “AI First”, aiming to put AI into every app, at the same time further developing the platform services AI (AI as a Service).
Big ambition, opportunity, clearly, but human is the knot, though the team made products have more than 1,000 people. According to him Minh teen who specializes in WHO is is the general problem of the company in Vietnam and in the world. This not only affects the speed of development, but also forcing businesses to change strategy.
According to statistics from recruitment platform CareerViet demand for engineers WHO with positions such as machine Learning, Processing natural language, computer vision... in the year 2025 has doubled compared to the same period 2024 and nearly 2.5 times higher than 2023. In it, the need for recruitment position management recorded an increase 5-6 times. “This reflects the businesses are not only looking for the real exam, but also need to team lead, to have the ability to build strategy and deployment of AI project large scale,” the analyst's assessment platform.
In the context of demand app development WHO rise strong, void of human resources also revealed. In many businesses, the shortage of people WHO do take place in all phases, from research, development to commissioning products, from enterprise scale to startup.
DR. Nguyen Tho Chapter, CTO of AI Or said with the peculiarities of business startups, they need the “full-stack” (capable of handling the entire floor of a software application) when developing AI feature. However, this is not simple because of the need to “build is thinking products for engineers WHO make technology serve users best in limited available resources to find people who really understand in depth about the models WHO have the ability to extend and optimal when products serve large user”.
According to him, on the market there are multiple engineers can do a good portion of the presentation or test for a feature, but the development of a real product stable operation at the scale of millions of users who left the challenge is completely different. Startup this maintained the progress of the project thanks to additional personnel rich experience mentoring and training young engineers now in the process of work.
Statistics of CareerViet shows, demand for personnel WHO not only limited in the technology industry, which has spread to other industries such as banking, finance, ecommerce, manufacturing, education... that the competition is increasing. The employers are willing to pay high salaries for engineers WHO, with levels range widely depending rank and experience, which, in average salary 11-20 million for students new to the field, 16-37 million for the basic staff, 25-38 million for the team and can be up to 40-60 million for management. In 2025, each recruitment engineer AI on the platform, attracting an average 600-1.000 views.
“This is testament to the level of attractive and competitive nature of the role of this” represents CareerViet reviews.
Journey fills a gap personnel WHO
From the enterprise perspective, according to mr. Bui Thanh Minh, who made WHO in Vietnam “are so lacking in both quantity and capacity” compared to demand and the development of the field of AI, especially after ChatGPT birth mark leap of SOMEONE giving birth.
From many years, Vietnam has built a team of engineers information technology strongly with the number of hundreds of thousands. However, the group specializes in WHO accounted for only few. According to experts, personnel and WHO IT in general need to have technical background and logical thinking the same, but the array WHO need more knowledge about math, data and machine learning to build systems that can “learn” and “predict”. “In other words, the hr WHO is the deeper of the IT employees, combining technical skills with thinking, research and data analysis”, leader of a unit study on artificial intelligence explains.
Analysis of the deficiency, DR. Le Quang Minh, Deputy director of information technology – vietnam National University, Hanoi, said, according to the survey results of the Institute, the training facility in Vietnam, with much difficulty, the largest of which is lack of foundation mathematics and programming, not enough high quality data to experimental and trouble when the access study material updates intensive. In addition, the training and practice in business “has not align with each other”, leading to the condition primarily student practice on the simulator project instead of joining in fact, from that limited application when out of school. This also leads to the idea, the possibility of building products, ONE of the kids in Vietnam often learn as foreign, leading to a latency of 3-5 years, equivalent to a course of training students.
To focus for AI, at some schools, the training program has been adjusted in the direction of strengthening background knowledge, increasing the number of credits in math, high to ensure students have enough mathematical prowess to grow and learn in depth model WHO.
From the business side, the deficiency also causes they are forced to adapt own way, as people have a good foundation and then more training, or transfer the software engineer good to ANYONE, although this process lasts monthly and high cost.
Mr. Minh said the business now implement a training program re-engineering IT tradition to do WHO, in collaboration with experts Vietnamese people are doing in the multinational technology company as big as Adobe, AWS, or Google, to solve complex engineering.
A number of corporations other technology is also going in the same direction when the combination of internal training and partnership with universities. The young engineers are involved in the project it's in environmental studies app helps shorten the gap between learning and production.
Prepare for the future
According to the director General CareerViet Hung Bui, recruitment trends engineer WHO in Viet Nam, next time will continue to grow, but the “focus on quality over quantity mere”. Two main trends are unit forecasts that demand shifted about ONE app, that business will prioritize search for the experienced engineer, deployment models WHO directly solve specific business instead of the new location began to learn the study. Monday, AI application in work is becoming criterion when recruiting in any industry, including traditional areas such as banking, retail, healthcare and manufacturing. “This is a great opportunity for all applicants who are interested in entering the market this dynamic,” mr. Hung said.
About training issues, the Deputy director of the Institute of information technology, Le Quang Minh recommendations state need the sync list to promote research, development and application of AI in Vietnam, including construction of the joint research program, establish a fund to support research, organize workshops to exchange expertise. In addition, the unit should have policies to develop, share, tapping database country/industry/local to the organization, the business can develop products WHO peculiarities of Vietnam.
“Should give priority to quality human resources training AI: the human resources development and human resources applications AI in work,” he stressed.
Report economy WHO Vietnam by NIC and Jica made to be published on march 4 reviews team of talents in the field of AI in Vietnam is growing strongly thanks to a focus on IT education. Until 2023, 165 university has implemented training programs in the IT industry, with the number of graduates increased by 4% per year. In the future, human resources of Vietnam's IT industry is expected to increase by 9% from 2022 to 2026, up 530.000 people.
The report workforce in the country still lack practical skills, in particular about the tools and framework WHO, due to the cooperation with business in the industry and the number of training programs, in-depth AI is still limited, however, have the advantage in background knowledge about math. “With the foundation of knowledge WHO aggregate and solid math, Vietnamese students have a competitive advantage in this area,” the report stated.
Source: vnexpress