AI application predicts the ability to pass university

In the announcement on the morning of June 19, AI developer Hay said that the feature had just been added to the application and website, after the high school graduation exam ended, preparing to enter the stage of announcing results and submitting wishes. According to the developer, the tool acts as an AI assistant to partly assist candidates in evaluating their chances of admission, finding suitable majors, and building a wish list based on personal data.

On the chatbot interface, users enter information, declare wishes and receive assessment results. Input data includes test scores, transcript scores, priority scores and results of competency assessments. From there, the system compares with the data set compiled by AI Hay from reputable sources to predict the likelihood of admission.

Results are displayed in natural conversation form, with explanations and suggestions for suitable schools and majors. Users can also continue to communicate with the chatbot to compare options and complete their wish list before officially registering.

According to the developer’s assessment, the college admission season in recent years is no longer a story of only comparing test scores with benchmark scores. With each major chosen, candidates must simultaneously handle many factors such as admission methods, criteria, tuition fees, employment trends, geographical location, personal capacity and suitability for the major. This makes the decision-making process more complicated than before, especially when data is scattered and can change continuously over each period.

“This is where AI tools come into play,” said Tran Quang Duc, CEO of AI Hay.

Predict with data

Last year, AI Hay also developed a tool to predict university admission rates, attracting more than 50,000 users, and bringing the application to the top 2 in the educational application rankings on the App Store in Vietnam. This is experience that helps them collect data from important sources, especially from universities, when the website structure and announcements of each school are different.

This year, the platform expanded to an “AI admissions assistant” model. The new feature combines benchmark data, school and industry information, targets, tuition fees, admission methods and career-oriented personality tests.

“The new point is not only in predicting pass or fail, but in the ability to help candidates understand why a request is suitable, where the risks lie and what additional options should be considered,” said a representative of the developer.

For candidates who have scores and have determined their aspirations, the chatbot will evaluate the possibility of admission based on comparing scores with recent years’ admission data. The system uses the benchmark scores of the last three years combined with factors such as admission criteria, admission methods, subject combinations and benchmark score fluctuation trends to make judgments.

The results are divided into several levels such as “comfortable parking”, “close to the mark” or “at risk of failing”. In addition to predicting the likelihood of admission, the chatbot also suggests adjusting aspirations, adding safe options or suggesting alternatives suitable to the candidates’ abilities.

For users who do not have scores but are interested in a specific school or industry, the chatbot provides information such as benchmark scores for years, admission quotas, tuition fees, training programs and admission methods. Meanwhile, candidates who have scores but have not yet determined their aspirations can receive suggestions on schools and majors that are suitable for their current score.

After the initial consultation, users can continue to learn more about each option. When requested by users, the system can support comparisons of study costs throughout the course, training quality, facilities, career opportunities as well as personal criteria such as geographical location or tuition level.

To ensure the accuracy of information, AI Hay said enrollment data is collected from official sources such as the Ministry of Education and Training, departments, university websites and press agencies. Data is updated regularly throughout the enrollment season, especially during periods when there are many changes in criteria, tuition fees or admission methods.

A new feature in this year’s admission season is the “Career Guidance Pot” quiz feature. The system will provide a 25-question multiple choice test, helping to classify users according to personality groups, thereby combining with actual test scores to suggest more suitable majors.

According to Mr. Tran Quang Duc, the goal of the tool is not for AI to decide for candidates, but to help users have more data and perspective before an important choice. According to him, this is a big decision and requires understanding, support, and discussion even from the candidate’s family, friends, and teachers. The role of the platform is to use AI technology to provide complete information, with clear origins, to help candidates standardize the admission suggestions effectively, in accordance with their own capabilities, financial conditions, career guidance.

“This is also the value that a purely Vietnamese AI aims for: bringing new technology to very Vietnamese problems,” Mr. Duc said.