Indonesia Elections Commission (KPU) Cuts Legal Research Time by 90% with Nutanix Enterprise AI

Title

Indonesia Elections Commission (KPU) Cuts Legal Research Time by 90% with Nutanix Enterprise AI

Content

Indonesia's General Elections Commission (KPU) deployed Nutanix Enterprise AI on Nutanix Cloud Infrastructure to build a Bahasa-language legal research assistant, cutting the time its legal team spends answering regulatory queries from 10-12 minutes per question to under 3 minutes — a 90 percent improvement — while consolidating decades of fragmented legal documents from the JDIH (Legal Documentation and Information Network) into a single searchable platform. Andre Putra Hermawan, Kepala Divisi Pusat Data dan Teknologi Informasi (PUSDATIN) at KPU, said: "AI cut our legal research time to under 3 minutes — boosting team satisfaction and ensuring instant access to accurate electoral records," describing the operational shift the new system enabled. The shift to AI-driven legal research addresses the structural challenge of running elections across the Indonesian archipelago of more than 210 million registered voters per cycle. Legal teams historically downloaded hundreds of pages and compared new regulations line by line against existing laws. Queries such as "What changed in the election funding regulations between 2022 and 2024?" or "Which provincial rules contradict national guidelines on campaign duration?" now return instant structured responses. The Nutanix architecture combines Unified Storage for the legal document repository, the Nutanix Database Service for full database visibility and protection, and the Nutanix Kubernetes Platform to host Llama 3 large language models. Nutanix Professional Services partnered with local application providers to deliver a turnkey AI chatbot tailored to KPU's workflow, with in-built access controls and audit trails aligned to local and international regulatory standards. Architecture spans Nutanix Cloud Infrastructure with Unified Storage for legal documents, Nutanix Database Service for database visibility and protection, and the Nutanix Kubernetes Platform hosting Llama 3 LLMs. The Bahasa-language conversational assistant includes structured response generation with relevant clauses and supporting screenshots. Operational impact at KPU includes a 90% reduction in legal research time, from 10-12 minutes per question to under 3 minutes, with decades of fragmented legal documents consolidated into a single searchable platform. The system covers presidential, legislative, and regional election legal workflows. Looking ahead, KPU is preparing to expose the conversational Bahasa assistant to Indonesian citizens, with the next phase covering migration of remaining electoral data types into the AI system and enhanced data security tooling ahead of the 2029 election cycle, positioning the deployment as a national reference for inclusive public services.