Start with one narrow use case – an internal knowledge assistant, a citizen-facing FAQ, document or email triage, structured transcription – ship it in a few short sprints, then expand. Avoid generic strategy decks and big-bang transformations. The most useful public-sector AI is unglamorous: it removes drudgery from staff and frees them for the work that needs human judgement.
Real AI government examples we have delivered – to name some:
Tannenhof Berlin-Brandenburg is an addiction therapy provider running a Civic Coding-funded AI transcription pilot with us. Therapists used to take paper notes and re-enter them into the Patfak digital patient file system. We built a prototype on EU-hosted infrastructure that records sessions on work iPhones and iPads, transcribes them and produces structured summaries formatted for German Pension Insurance (DRV) reporting. We started with a premortem on 5 June 2025 that mapped failure modes including legal change, data security and user acceptance. We then A/B-tested against the standard Patnova product and renamed UI terminology from "Monologue" to "Einzel" after clinical-team feedback. The pilot used 32 hours of Civic Coding consulting.
Kompetenzz runs a production retrieval-augmented generation (RAG) chatbot inside their HumHub social network for 1,000+ members, on a stack of n8n in Berlin, Qdrant in the EU and GPT-4 via Microsoft EU sovereignty, delivered in four sprints. GDV (German Insurers Association) runs the same architecture across tens of thousands of policy documents for 400+ insurance companies. innatura runs an AI email agent in pilot that classifies enquiries and drafts replies for human review. Under Civic Coding, N3XTCODER has delivered AI consultation across 100 social-impact projects.
Example: AI Knowledge Assistant for policy documents
By playing the video, you consent to the transfer of data to the respective video platform (e.g., YouTube, Vimeo).
Grounded in your own data
EU-hosted infrastructure
Human-in-the-loop by default
Real systems in production
Kompetenzz (1,000+ members), GDV (400+ insurance companies), a leading German association, innatura, Tannenhof
Staff freed from drudgery
from triage, documentation and search to focus on judgement work
Time to first project
first working version delivered in weeks – the Kompetenzz timeline
Compliance built in
EU AI Act, GDPR and your internal governance designed in from the first sprint
Maintained by your team
low-code architecture and documentation so non-technical staff can operate it
1. Pick a narrow use case
2. Build a working prototype
3. Hand over and expand
We bring a decade of impact-tech experience and more than 160 AI projects since 2019. Through our free AI for Impact course, more than 100,000 people have learned how to use AI for the common good. We do not run inspiration days. We run scoping sessions and build engagements that ship, the way we have delivered AI for the organisations below:
Big-bang AI transformations almost always fail. Pick one narrow use case, ship it, then expand. We are wary of any RFP that asks for an enterprise-wide AI strategy in one tender.
Some processes should not be automated. Where the cost of a wrong answer cannot be tolerated even with human review, the honest answer is "not now".
If your data is not ready, no AI will fix it. A smaller data project is usually the right next step before any AI work.
Where should we start with AI in the public sector?
Is AI in the public sector compatible with EU rules?
We don't have AI expertise in-house. What do we do?
How long does a first public sector AI project take?
Can N3XTCODER work under public procurement?
What are some real AI government examples?
Tell us about your organisation and the challenge. We will reply with a proposal and a date.
Simon Stegemann
Co-Founder and CEO
