The OECD.AI Policy Navigator

Our policy navigator is a living repository from more than 80 jurisdictions and organisations. Use the filters to browse initiatives and find what you are looking for.

AI Assistant for Automated Generation of Indicator Fact Sheets


-
Added by:   OECD analyst
Added on:   27 Oct 2026
Updated by:   OECD analyst
Updated on:   27 Oct 2026

The AI Assistant, combining a custom GPT with Python, automates the creation of 16-field indicator fact sheets, scaling from a 2025 pilot with 65 local documents to a national-level tool for 2025-2026. The updated system covers the Multi-sectoral Strategic Plan, the Coordinated Development Plan and the Institutional Strategic Plan (PEI), incorporating semantic validation against 2024 guidelines, intelligent navigation and batch processing to enhance efficiency across all levels of government

Name in original language

Asistente IA para Generación Automatizada de Fichas Técnicas de Indicadores

Initiative overview

The National Center for Strategic Planning (CEPLAN) development and evolution of a custom AI-based assistant (Custom GPT) designed to automate the creation of technical sheets for strategic indicators. Originally conceived as a pilot solution for local governments, its successful implementation in 2025 led to strategic nationwide scaling. The project serves as a cornerstone for Peru’s national planning system, ensuring that indicators across all levels of government meet rigorous technical standards with significantly reduced manual effort.

Phase 1: Focused and Validated Solution (Completed in 2025)

The first phase was designed to address a specific bottleneck: providing technical assistance to local governments in drafting technical sheets. The primary goal was to automate the generation of these sheets based on the Institutional Strategic Plan (PEI) and the Institutional Operational Plan (POI). The pilot was validated through a real-world case study, resulting in the successful creation of 65 standardised technical sheets.

The solution integrated a Custom GPT with a Python script. Using the python-docx library, the system replaced placeholders in a Word template to generate the final official document.

Phase 2: National Scaling and Expansion (In Progress, 2025–2026)

Following the positive results of Phase 1, CEPLAN identified a natural opportunity to expand the tool's scope. This evolution is not a new project but a strategic expansion of the existing product to cover all instruments of the National Strategic Planning System (SINAPLAN).

Expanded Scope

The assistant now encompasses a broader range of strategic instruments:

  • PESEM (Multi-year Sector Strategic Plan): Covers Sector Strategic Objectives (OES) and Actions (AES).
  • PDC (Coordinated Development Plan): Includes objectives and actions at the regional level (PDRC: OER/AER) and local level (PDLC: Province and District).
  • PEI (Institutional Strategic Plan): Now available for all public entities nationwide, moving beyond the initial local government focus.

Key Features of the Scaled Version

  • Intelligent Navigation Tree: The assistant features a structured menu that adapts dynamically based on the specific planning instrument (PESEM, PDC or PEI), the level of government (National, Regional or Local) and the type of element required.
  • Semantic and Structural Validation: To ensure high data quality, the system validates indicator names and descriptions against the Guide for the Preparation of Indicators CEPLAN 2024 before processing begins.

Other relevant details

Results, outcomes and impacts: The AI initiative by CEPLAN offers transformative benefits for Peru’s National Strategic Planning System. Institutional benefits include national standardisation, which ensures uniform technical quality across all levels of government; extreme efficiency, reducing drafting time by 90–95%, from 50 minutes to under five minutes per sheet; massive scalability, enabling simultaneous high-quality support for 26 regions and over 1,800 local governments; and strategic focus, freeing specialists from repetitive tasks to focus on high-value analysis. User benefits include democratised access, enabling small local teams to achieve the same technical rigour as large ministries; and error reduction, minimising technical observations in submitted plans. Challenges and lessons learned: The AI initiative has addressed the following challenges: AI precision, involving the need to ensure technical consistency in generated justifications and baseline values; semantic complexity, involving the need to balance automated corrections of poorly drafted titles with the user’s original intent; exception handling, requiring the development of specialised logic for atypical indicator structures, such as composite indices or specific ratios; adoption and transparency, requiring the management of user expectations so that generated values are understood as referential rather than actual data; and regulatory alignment, requiring the assistant to be kept updated with frequent changes in CEPLAN’s methodological guidelines. Lessons learned include the importance of real-world validation: testing the tool on a 65-sheet pilot for local governments provided the evidence needed to justify national scaling. They also include the role of phased evolution in reducing risk: starting with a focused pilot (Phase 1) allowed for refinement before a complex nationwide expansion (Phase 2).

About the policy initiative


Category:

  • AI policy initiatives, programmes and projects

Initiative type:

  • AI use cases/projects in the public sector

Status:

  • Active

Start Year:

  • 2025

Target Sectors:


OECD AI Principles:

—

Other relevant urls:

—