Faster procedures: how MIA, the State's new artificial intelligence, works

The national government introduced MIA, the first artificial intelligence agent of the Argentine State, with the promise of streamlining procedures and optimizing public resources.

The digital tool aims to reduce thousands of hours of waiting and provide a faster, more scalable service for citizens.

From TINA to MIA: the evolution of the digital State

MIA is an evolution of TINA, the official chatbot that already allows users to manage more than 600 online procedures. TINA processes an average of 1.5 million conversations per month and has become a key channel of contact with the State.

With MIA, generative AI capabilities are added, making the service more agile and personalized, reducing the workload of support teams.

Services MIA offers in this first stage

It will initially focus on entrepreneurs and SMEs, offering guidance on business formation, exports, and key information for productive development.

Ahora se puede hacer de forma 100% digital

It also integrates with the Mi Argentina app, which has more than 26 million users, and seeks to ease the help desk that receives about 80,000 inquiries per month.

Technology behind the new agent

MIA was developed on the Meta Llama 4 model, an open-source system that provides speed, low operating cost, and scalability. This allows for savings of up to ten times compared to other high-performance models.

According to Alejandro Zuzenberg from Botmaker, the project prioritizes privacy and data auditing, making it an example of responsible innovation in the region.

MIA fue desarrollado sobre el modelo Llama 4 de Meta

Security and reliability of information

A key point is that MIA is fed by a closed knowledge base, built with official procedures and services. This ensures that responses are reliable and secure.

The controls implemented allow data to be managed in local and auditable environments, strengthening the protection of public information.