{"slug":"finnish-authority-builds-ai-apuri-a-rag-chatbot-platform-where-civil-servants-create-team-chatbots-with-12-chatbots-in-production","url":"https://findausecase.com/use-cases/finnish-authority-builds-ai-apuri-a-rag-chatbot-platform-where-civil-servants-create-team-chatbots-with-12-chatbots-in-production","title":"Finnish authority builds AI-apuri, a RAG chatbot platform where civil servants create team chatbots, with 12 chatbots in production","description":"A Finnish authority introduced AI-apuri, a commercial platform on which civil servants build retrieval-augmented chatbots for their own teams, with inference residency controlled and training data capture not permitted. The source reports 12 chatbots in production and over 60 explored use cases, with estimated savings of about 30 minutes per drafted reply and up to 15 minutes per consumer complaint translated and formalised. A human reviews every final communication.","company":"Finnish public authority (not named in source)","industry":"Government & Public Sector","country":"Finland","aiCapabilities":["Generative AI"],"technology":["AI-apuri"],"deployment":"Unknown","problemStatement":"Generic large language models answer from general training data and cannot reliably rely on curated, task-specific information, while many work processes at the authority require up-to-date internal guidance, legal interpretations and structured external sources.","solutionApproach":"AI-apuri is a commercial platform that uses a retrieval-augmented generation architecture, where chatbots retrieve answers from defined internal and external data sources. Civil servants build chatbots for their own teams with support from the ICT unit, can instruct them to answer only from curated sources, and a human reviews and sends the final communication. Inference residency is controlled and training data capture is not permitted.","businessValue":"The source reports 12 chatbots in production and over 60 explored use cases. It says drafting a reply may save about 30 minutes per case and translating and formalising a consumer complaint up to 15 minutes; impact is measured through usage data, funnel progression and case-level time saving estimates.","evidence":{"band":"high"},"sourceUrl":"https://oecd.ai/en/dashboards/policy-initiatives/ai-helper","dates":{"publishedAt":"2026-11-15T05:48:55.915Z","publishedAtSource":"pipeline","updatedAt":"2026-11-15T05:48:55.915Z"},"license":"Open for reading and citing with a link to https://findausecase.com/use-cases/finnish-authority-builds-ai-apuri-a-rag-chatbot-platform-where-civil-servants-create-team-chatbots-with-12-chatbots-in-production. Bulk republication requires permission."}