Enfarm Agritech, an artificial intelligence platform that supports farmers in making farming decisions, surpassed four startups to win the AI ​​Startup Award 2026.

The information was announced by Mr. Nguyen Do Dung, co-founder and CEO of Enfarm, on November 2. The award was awarded within the framework of the 19th German-Asia Pacific Business Conference (APK 2026), taking place in Seoul, Korea. This is a high-level forum connecting the German business community with the Asia-Pacific region, gathering about 800 business leaders, policymakers and industry representatives.

The AI ​​Startup Award final round has 5 representatives selected from thousands of startups across Asia, from Vietnam, Korea, Singapore, India and China.

Teams have three minutes to present their solutions, then answer questions directly from three senior leaders of the German Government and businesses, including Deputy Minister of Economy and Energy Stefan Rouenhoff, General Director of B. Braun Group Anna Maria Braun and President and General Director of Siemens Roland Busch.

The winner is decided by direct voting by delegates at the hall, including business leaders and German Government officials. Enfarm won the highest number of votes.

At the competition, Enfarm introduced Famos, a digital agronomy platform using artificial intelligence, helping farmers understand the status of farming and receive specific recommendations on what to do, when to do it and why.

This foundation was developed from the fact that each farm has different conditions in terms of soil, weather, plant variety, growth stage and farming history. Therefore, for the same plant but the way of care, watering or fertilization cannot be applied to a general formula.

Famos combines multiple layers of data on soil, weather, crops, growth, farming history and user observations to provide guidance tailored to each farm.

Inside the platform is Agos, a specialized AI tool for agriculture developed by Enfarm. Unlike general-purpose chatbots, Agos is designed to focus on agronomic decision support.

A notable point is that the system does not let the large language model create important parameters such as fertilizer dosage on its own. Quantitative data is processed by specialized models, then recommended to be checked according to crop, soil, weather and safety rules before being sent to users. When data is lacking, the system is designed not to automatically generate information to fill the gap.

According to the business, this approach is to limit risks when applying AI to production. An incorrect recommendation for fertilizer amount or timing of care can affect yield, even causing damage to the entire crop.

Enfarm tested Agos with more than 1,000 questions about agriculture, comparing it with three popular AI platforms in the world. The four systems are provided with input data, farm context and knowledge sources. As a result, Agos achieved the highest score in about 97% of the questions, and also had the lowest rate of answers flagged for violating safety regulations in the test group.

One of the bases for Enfarm to develop Famos and Agos is a smart soil sensor system that has been deployed directly on farms.

In two years, the business put more than 2,200 sensors into the field. About 90% of customers can install the device themselves without needing a technician to assist. The implementation process helps Enfarm accumulate data and experience on soil characteristics, crops, farming behavior as well as how to translate the collected information into useful recommendations for farmers.

Famos is designed to work even when users do not have sensors, phone signal or internet connection. The goal is to bring personalized agronomic knowledge to producers in remote areas at low cost.

Enfarm Agritech is an agricultural technology enterprise established in Vietnam, operating in the fields of soil sensors, Internet of Things (IoT), agricultural software and specialized artificial intelligence. Before the award in Seoul, the business was also selected by many domestic and foreign organizations and forums as a typical AI project.

Le Tuyet