dinero_admin 
The Agencia Nacional de Inteligencia Artificial de El Salvador (ANIA) announced the creation of Nemotron-Personas-El-Salvador, the first open dataset of synthetic people developed specifically to represent the country’s demographic reality.
The initiative marks an important step in building national capabilities for the development of artificial intelligence adapted to the salvadoran context.
According to ANIA, this dataset was developed in collaboration with NVIDIA and WideLabs, a company specializing in sovereign artificial intelligence in Latin America. Its purpose is to provide a secure and representative information base that allows for the design, training, and evaluation of AI systems without using the personal data of real citizens.
These so-called “synthetic people” are profiles generated using artificial intelligence that reproduce real demographic characteristics of the population, such as age, geographic location, occupation, and socioeconomic conditions. Although they are based on official statistics, they do not correspond to existing individuals, which guarantees privacy protection and eliminates risks related to the use of personal information.
According to ANIA, Nemotron-Personas-El-Salvador was built using statistical distributions derived from official sources in the country, including information from the Censo de Población y Vivienda 2024. This allows the profiles to more accurately reflect the diversity of salvadoran society and its cultural, social, and economic characteristics.
The organization emphasized that the project addresses a growing need within the technology industry. As artificial intelligence evolves from simple conversational assistants to agents capable of autonomously performing complex tasks, it becomes essential to have data that accurately reflects the local realities where these tools will be used.
One of the main benefits of the new dataset is that it will facilitate the development of digital assistants and automated services geared toward citizen services. With nearly one million synthetic profiles representing the population of the country’s 14 departments, developers, researchers, and institutions will be able to test technological solutions in scenarios that more closely reflect the national reality.
The Agencia Nacional de Inteligencia Artificial de El Salvador (ANIA) emphasized that privacy is one of the project’s cornerstones. All records included in Nemotron-People-El-Salvador are entirely artificial and contain no information that could identify real individuals. This feature allows the dataset to be used immediately in research, testing, and technological development without compromising the protection of personal data.
Another key aspect is that the resource will be available under an open license, allowing universities, research centers, startups, technology companies, and government institutions to access it free of charge. This aims to foster innovation and accelerate the development of artificial intelligence-based solutions within the country.
The launch of Nemotron-People-El-Salvador also positions the country within an international initiative driven by NVIDIA, in which various nations developing data infrastructures to strengthen their artificial intelligence capabilities participate. El Salvador’s inclusion in this program reflects its interest in building its own technological tools and reducing its dependence on models designed for other markets and cultural contexts.
For ANIA, the availability of this dataset represents a strategic foundation for future national technological development. The institution believes that having resources adapted to the Salvadoran context will allow for the creation of more precise, useful, and aligned artificial intelligence systems, better suited to the needs of the population, while simultaneously strengthening the country’s innovation ecosystem.
With this launch, El Salvador takes another step forward in its digital transformation strategy and positions itself among the countries in the region seeking to develop their own artificial intelligence capabilities, using local data and models designed to meet their specific needs.
You can also read:
