Agento: the AI from Eivor Systems that brings public administration closer to the citizen
Eivor Systems, a technology company based in the Canary Islands specializing in software, automation, and artificial intelligence, has developed
Agento, a conversational assistant designed for public administration. Its goal is for citizens to be able to ask in natural language, for example, what they need to apply for assistance or how to register, and receive a clear response without having to navigate through websites, PDFs, and electronic offices.
The system is based on an information retrieval architecture (RAG): instead of responding with generic knowledge, it consults the official documentation of each agency and generates responses tailored to that source. According to
Oliver Martín Ramírez, CEO of
Eivor Systems, this difference is key in the public sector, where "it is not enough for something to sound reasonable"; the response must be based on updated institutional information.
One of the strengths of
Agento is its ability to recognize when it does not have enough information, directing citizens in those cases to an official source or a human support channel, to reduce the risk of errors or "hallucinations" typical of generative AI. Additionally, it includes an analysis panel that allows the institution to detect recurring questions and possible problems in its own communication.
From
Eivor Systems, they insist that the tool does not replace public staff or participate in administrative decisions: its function is to provide initial guidance and resolve repetitive inquiries, freeing up professionals' time for cases that require human judgment.
Agento is prepared to operate in Spanish, co-official languages, and other languages, and considers deployments on infrastructure located in Spain or the European Union, with privacy integrated from the design. In line with
LICAI, its product for public procurement, Eivor Systems presents Agento as a scalable solution, capable of growing both within and outside the Canary Islands, adapting to institutions of any size.
