Handling applications and cases.
AI checks whether a file is complete, applies the relevant policy and prepares the case. The case officer reviews it and takes the decision.
AI for government that residents can trust.
Faster public services. Decisions people can still explain.
Public organisations handle more requests, cases and rules every year, often with the same number of people. AI can take on part of that work, but only if every step is lawful, explainable and secure.
Capisoft helps municipalities, provinces, executive agencies and ministries put AI to work in a way they can account for: to residents, to their council and to the regulator.
Leave your details and we will contact you within one working day. A few short questions help us come prepared.
Public work runs on trust.
A municipality cannot choose its customers. Every resident has to be helped, every decision has to follow the rules and every outcome may have to be explained later, in an objection, a Woo request or a council debate.
That is why AI in government needs more than a good model. It needs clear boundaries: what the system may do, which data it may use and where a person decides.
We design those boundaries first and build the software around them.
Where AI helps public organisations.
Finding the right rules.
Staff ask questions about regulations, local policy and earlier decisions, and every answer shows the source it came from.
Answering residents and businesses.
Incoming questions are understood, answered where the answer is clear and passed to the right team when it is not.
Modernising older systems.
We help your developers understand, document and gradually replace legacy software, with tests and review on every change.
Built to fit the rules you work under.
Every public organisation answers to a framework of laws and standards. We take it as the design brief, not as paperwork at the end. Together with you, we work through:
- 01
AVG (GDPR): a clear legal basis, data minimisation and a DPIA where the processing requires one.
- 02
EU AI Act: the risk class of each system, human oversight and the documentation that goes with it.
- 03
BIO: security measures that fit the Baseline Informatiebeveiliging Overheid and your own information security policy.
- 04
Algorithm register: a plain-language description of each algorithm, ready to publish.
- 05
Fundamental rights: an IAMA (human rights and algorithms impact assessment) for systems that affect residents.
- 06
Accessibility: interfaces that meet WCAG 2.2, as required for government websites and apps.
- 07
Data location: agreements on where data is stored and processed, and which AI models may see it.
Capisoft is ISO 27001 certified for information security.
Every outcome can be explained.
Residents have a right to know how a decision about them was made. So we build systems that record what the AI did, which information it used and who approved the result.
AI supports the case officer. It does not replace their judgement on matters that affect people's lives, income or rights.
We also keep the description of each system up to date, so publishing it in the algorithm register is a matter of review, not a separate project.
If a simpler, rule-based solution does the job better, we recommend that instead.
From a first question to a working service.
- 01
Assess the task
We choose one process with a clear public value, map the rules that apply and assess the risks: privacy, fundamental rights and the EU AI Act risk class.
- 02
Test with real cases
We build a first version and test it on real, properly protected cases with the people who do the work, before any resident is affected.
- 03
Fit your procurement
We work within your purchasing rules, whether that is a small assignment, a framework agreement or a public tender, and supply the security and privacy documentation you need.
- 04
Run and improve
Once it is live, we monitor quality, keep the documentation current and improve the system as rules and needs change.
Which public task should be easier?
Tell us which process takes too much time today, which rules apply and what a better service would mean for residents and staff.
We’ll help you assess whether AI is the right answer, what the risks are and what a responsible first step would look like.