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.

Minute


Added by:   OECD analyst
Added on:   16 Nov 2026
Updated by:   OECD analyst
Updated on:   16 Nov 2026

Using a transcription engine and a large language model (LLM), Minute produces bespoke summaries in the formats that public servants really need. Users have the choice between uploading existing audio and video files, or recording live directly in the app.

Initiative overview

Local authorities deliver essential services to vulnerable residents through thousands of face-to-face interactions daily. Yet officers in these meetings face an impossible choice: be fully present with the person in front of them or ensure accurate documentation for case files. Traditional note-taking steals attention during the very moments when human connection matters most. This isn't just inefficient; it fundamentally undermines the trust and rapport essential to effective public service. After meetings, officers spend additional hours reconstructing conversations from fragmented notes.

Minute creates the space for genuine human connection by removing the burden of documentation during meetings. The tool automatically transcribes meetings up to OFFICIAL-SENSITIVE classification, freeing front-line workers to be fully present with residents. Operating in the background on secure infrastructure, it identifies speakers, making it ideal for multi-speaker settings, captures complete discussions, and creates transcripts in flexible formats, allowing officers to focus entirely on the human interaction.

In 2025, the tool was tested and consequently deployed across 22 local authorities as well as Ministerial Private Offices. Officers reported satisfaction in being able to focus on the person in front of them rather than note-taking. The code has been open-sourced on GitHub, enabling other organisations to adapt the solution.

Results, outcomes and impacts

All 22 local authorities that participated in the 2025 trial have deployed the tool owing to officers' reported satisfaction in enabling full attention during critical conversations with vulnerable residents. Multiple Ministerial Private Offices have adopted the tool. Code has also been published on GitHub to enable organisations to create customised versions for their specific needs, including Justice Transcribe, the Ministry of Justice's adaptation for transcriptions by probation officers.

Challenges and lessons learned

The 100% trial-to-deployment rate proves that solving genuine pain points and daily frustrations drives adoption far more effectively than top-down mandates. Documentation burden is universal across government, but implementation requirements vary. Open-source deployment allows organisations to adapt the tool to their own needs and nuances. It also helps accelerate delivery of solutions, while reducing duplication of effort. Leading with humanity at the centre, rather than pure productivity, can unlock genuine enthusiasm.

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:

—