AI to empower people in need
and those who help them
The first open-source AI agent platform built for public services.
Toda persona que necesite ayuda
pueda recibirla.
Una IA segura, eficaz y soberana para el servicio público
1Segura
Sigue tus protocolos en lugar de los reflejos del modelo, cita sus fuentes línea por línea, y tus expertos la validan antes de ponerla en servicio.
2Eficaz
Se mide en producción, con resultados sobre el terreno: tiempo devuelto a los equipos, respuestas más rápidas a los usuarios, un cumplimiento seguido en el tiempo.
3Soberana
Funciona en tu infraestructura, con tus datos, en código abierto. Puedes retomar el control en cualquier momento.
La IA genérica no se diseñó para los servicios públicos y sociales
Millones de personas dependen cada día de profesionales sobre el terreno. Pero la IA comercial no se diseñó para ellos.
No conoce sus protocolos.
No protege los datos sensibles.
Alucina cuando más importa.
La principal organización sin ánimo de lucro que aprovecha la inteligencia artificial para reinventar los servicios públicos a gran escala
Fundada en Silicon Valley en 2014, nuestras soluciones de IA de código abierto cuentan con la confianza de gobiernos, organizaciones internacionales y comunidades locales en ocho países, al servicio de millones de ciudadanos.
Saber másIntroducing the Bayes Platform
The first open-source AI agent platform built for public services. Here is what it does, in the real product.
The agent answers from your handbooks, protocols and official references: resource libraries, tagged documents, PDF import, a site indexed by crawl. Never from the open web.
The sources panel shows the document title and the extract used, so whoever reads the answer can check it. An answer nobody can verify has no place in a public service.
The agent's settings are your organisation's. Every reply has to report its own bookkeeping: session title, categories drawn from your list, sources cited. The platform enforces that when the model skips it.
Every change writes a draft. Publishing makes it the version that runs, named and dated in the history, and every reply carries its version badge. Nothing reaches the public that a human has not published.
Your frontline teams create and tune their own agents, with no code and no internal data team. Your IT keeps the path to production without being the bottleneck.
Evaluation campaigns replay conversation and extraction datasets against a chosen version. Reviewers rate the answers blind, an LLM judge gives the scale, and two runs compare side by side.
Every exchange is filed under the life situations you defined. The topics your answers cover badly surface on their own, instead of waiting for a complaint.
Each agent exposes itself as an MCP server, so the assistant or portal your institution already runs calls it as a tool. A widget on your site or a public page when that is simpler. Nobody changes screen.
Managed cloud, or your own infrastructure from a single Docker image. Document isolation is written into the retrieval query itself rather than a checkbox, the retention period is yours, and the code is MIT-licensed.
AI grounded in your own protocols.
A general-purpose assistant answers plausibly but sometimes wrongly; grounded in your official, cited guidelines, it answers correctly.
DemoThe same clinical question, put at once to a generic assistant and to the MCP-verified platform.
A proactive AI work coach.
An agent that draws on the platform tools to answer, and stays within its remit: out of scope, it routes you to a human counsellor.
DemoA coach that calls tools, surfaces openings and benefits, and hands off to a counsellor when a request is out of scope.
A whole public service, by chat.
A full administrative procedure handled end-to-end on WhatsApp, without leaving the conversation.
DemoA diploma request: identity (e-ID), auto-fill from the official records, and payment.
En la prensa
Una selección de artículos de prensa sobre nuestra misión y nuestro trabajo.