Developing an automated and efficient validation & publishing pipeline in GitLab
Zorginstituut Nederland
Data & AI for care for older people
A lot of data is recorded in care for older people, often for someone else.
Care records, staff rosters, hours worked, client feedback and accountability data for regional care offices, the inspectorate and sector bodies are spread across the electronic care record (ECD), the rostering system and separate reports. The same figures are collected several times, while individual care locations only see afterwards where staffing falls short of care needs. Data and AI bring this information together, make reporting repeatable and give care home and community care teams the insights they need.
Challenges,
Care homes, community nursing teams and home care providers face a growing older population and a tight labour market. Care staff spend a substantial part of their shifts on record-keeping, while organisations submit quality and operational data to several parties at once. This includes health data about vulnerable people, which is sensitive personal data. It calls for one reliable source, clear rules on who can see what, and tools that support care staff rather than create more work.
How we support you,
We bring data from the ECD, staff rosters and financial records together in one data foundation. We use it to build reports on staffing, resource allocation by location and the accountability data you need to submit. For KIK-V, a programme in which organisations across the care home sector are streamlining the exchange of accountability data, we built an automated validation and publication pipeline for Zorginstituut Nederland. We also help care staff find the right information in records and protocols more quickly. These are just some examples of what is possible. The best solution always depends on the challenge and the organisation.
DSL in numbers,
40
Data experts
10+
Years of experience
200+
Projects completed
8,7
Client satisfaction
Get in touch with our experts,
Want to know what we can do for your organisation?
Younes Seghrouchni
Managing Consultant GenAI & Data StrategyE-mail info@datasciencelab.nl
Carmen Wolvius
Managing Consultant Data Science & AIE-mail info@datasciencelab.nl
Stef Stoelwinder
Business DevelopmentE-mail info@datasciencelab.nl
Where we do that,
Latest blog posts,
Read our latest insights here.
Three lessons from our AI projects
Over two years, we supported dozens of AI projects, from initial exploration to a model that runs every night. Some went exactly as planned; others…
From pilot to production: why most AI projects stall
Many organisations have now run an AI pilot. The model works, the demo gets applause, and then nothing else happens. In our experience, most projects…
Responsible AI: from principles to practice
Artificial Intelligence is developing rapidly. New models appear almost every week, and more and more organisations are experimenting with AI. At the…
The organisations that do this,
Frequently asked questions,
AI can summarise reports in care records, compare roster data with care needs and check accountability data before it is submitted. This leaves care staff with more time for residents and clients.
Start with a question that keeps coming up, such as how staffing at each location compares with care needs, or how much work it takes to submit accountability data. Using data you already have, you can quickly see what a dashboard or model would add to your current way of working.
More than you might think: care records and reports in the ECD, rosters and hours worked, client feedback, incident reports and financial records. The work lies in connecting these sources by location and team, so they provide one reliable picture that you can also use for reporting.
In the KIK-V programme, organisations across the care home sector work together to streamline the exchange of accountability data between those requesting information and care providers. For Zorginstituut Nederland, we developed CI/CD pipelines in GitLab to ensure controlled releases of KIK-V products. Read the case study about the validation and publication pipeline.
We determine what data is genuinely needed for each question, use aggregated or pseudonymised data where possible, and manage access by role. A model provides an alert or a summary; the care worker or specialist in elderly care medicine decides what to do with it. We have written separate pages for other care sectors about data and AI in healthcare and care for people with disabilities.