evidoria

← Back to browse

Good practice Imported

deliberAIde — AI-Assisted Participatory Planning Conferences in Siegen-Wittgenstein

Germany · Siegen · See the Germany profile

Evidence: Descriptive / self-reported Top 76% 47/100 · Ask Evidence Copilot about this practice

Germany's most forested district piloted deliberAIde, an AI platform that transcribes, anonymises and clusters citizen input at social-planning conferences, helping ten municipalities turn table discussions into concrete proposals for local politicians in 2024.

10 conferences
Participatory planning conferences held (2024)
30–90 attendees
Attendees per conference (2024)
2 proposals
Citizen-driven proposals advanced to local politicians (2024)
deliberAIde — AI-Assisted Participatory Planning Conferences in Siegen-Wittgenstein

Details

Maturity
Pilot
Promoter
Kreis Siegen-Wittgenstein, with deliberAIde and the University of Siegen
Period
2023–2024
Region (NUTS)
DEA59
Keywords
social planning, civic tech, deliberative democracy, speech-to-text AI

Context

Kreis Siegen-Wittgenstein, a district of about 270,000 residents in North Rhine-Westphalia where 71% of the land is forested, runs annual "Sozialplanungskonferenzen" (participatory social-planning conferences) to gather residents' priorities for local social policy. In 2024 the district piloted deliberAIde, an AI platform built by an independent civic-tech team, as its second iteration after a 2023 pilot.

Activities

During ten conferences held across ten of the district's eleven municipalities, with 30 to 90 attendees each, deliberAIde was used at 10-12 discussion tables to transcribe speech in real time with automatic anonymisation, cluster similar ideas and themes, tag and categorise contributions, and generate summary reports and visual "digital sticky notes" of each table's discussion. The platform uses large language models and image-generation models for summarisation, visualisation and analysis, while human moderators remained central to facilitation — the AI's stated purpose was to free moderators to focus on dialogue quality rather than note-taking, turning documentation that previously took days into work completed in minutes.

Results

Two citizen-driven proposals from the 2024 conferences advanced to local politicians: a multi-channel AI platform to improve access to social support services, and a "third places" framework for rural childcare and intergenerational meeting spaces. The pilot ran at no direct cost to the district, which served as an alpha-phase testing partner; the University of Siegen supported evaluation of the social-support platform proposal, and the pilot was backed by the Friedrich Naumann Foundation's Global Innovation Hub in Taipei.

Implementation

Indicative cost
Low (< €50k) — The pilot ran at no direct cost to Kreis Siegen-Wittgenstein, which served as an alpha-phase testing partner for the deliberAIde platform.
Time to results
Short (< 1 year) — Second iteration in 2024 following an initial 2023 pilot; ten conferences held across ten of the district's eleven municipalities during 2024.
Staffing & skills
deliberAIde — an independent civic-tech team — built and operates the platform, Human moderators from Kreis Siegen-Wittgenstein facilitate each discussion table, University of Siegen supported evaluation of the social-support platform proposal

Conditions for success

  • Human moderators remain central to facilitation while AI handles transcription/documentation, freeing them to focus on dialogue quality
  • Real-time automatic anonymisation built into the transcription pipeline
  • Backing from the Friedrich Naumann Foundation's Global Innovation Hub, with the district serving as an alpha-phase testing partner

Commonly funded by

Philanthropic / foundation funding CERV — Citizens, Equality, Rights and Values

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful