Data engineering consultancy in Groningen
We connect, structure and standardise your data sources in a single data platform, as we did for the Concertgebouworkest with Databricks and dbt.
For Koninklijk Concertgebouworkest, we brought scattered sources together in one data platform where information is easy to find
The orchestra’s rich history was spread across systems, archives and data sources. By connecting, structuring and standardising data from multiple systems, we’re making information easier to find and widely accessible […].
Why do we still spend hours looking for data that’s already somewhere in our systems?
A record-keeping system, an archive, a few exports and a spreadsheet maintained by one colleague. When no one knows where to find what, every report or model starts with a search. A data platform with a catalogue solves that.
Sources connected, data standardised, and a catalogue showing you what’s where
Connect
Systems, archives and exports feed into a single platform, as when we connected data from multiple systems for the Concertgebouworkest.
Standardise
The same concepts, codes and dates across every source, recorded in transformations you can trace step by step.
Make data easy to find
A catalogue listing an owner, a description and access permissions for each table, so colleagues can find what they need themselves.
Keep it running
Loading steps with checks and alerts, so you spot a broken source before a report shows incorrect data.
We start with the source that takes the most time to search
-
Source inventory
Which systems, archives and exports exist, who manages them and which questions are still unanswered.
-
One source, end to end
From retrieving the data to a well-structured table with a description, checked by the colleagues who currently search through it.
-
Expand and hand over
We add sources one by one, with management handled by us or handed over to your own team.
Does an organisation in the north need its own data platform, or would something smaller do?
A good fit if
Data is spread across multiple systems and archives, reports and models each use their own copy, and no one can say which version is correct.
Not a good fit if
Everything is in one system with good reporting, or you only need a one-off analysis. In that case, a platform creates more management work than value, and we’ll tell you so.
What determines the scope
The number of sources, how old and messy the data is, who is allowed to see which data, the cloud you already use and who will take over management.
Can a team from Amsterdam work remotely too?
Yes. Most of the build work takes place in your cloud environment and can be done online. We come to Groningen for the initial assessment and testing with users.
A data warehouse, a lakehouse or just a catalogue to start with?
A data warehouse
Choose this when you mainly need to bring together structured data from a few systems for reporting. Read more about the data warehouse.
A lakehouse
Choose this when you also need to include documents, archives and large files, and models use the same data. If you have many types of sources, we recommend starting here.
Start with a catalogue
Choose this when the data is already there but no one can find it. A description and an owner for each source are the quickest way to make progress.
We've been building data platforms since 2016 as a Databricks and Microsoft partner
Data Science Lab works from Westerdoksdijk in Amsterdam, including for organisations far beyond the city. For the Concertgebouworkest, we built a data platform using Databricks, Delta Lake, Unity Catalog and dbt.

For an initial conversation, a list of your systems and the questions you need answered is enough
Stef Stoelwinder
Business DevelopmentStef is your first point of contact. In our initial conversation, we’ll discuss your data sources, what your colleagues currently spend the most time looking for, and which cloud you already use.
E-mail info@datasciencelab.nl
Phone +31 (0)20 3636 724
Which source currently takes you the most time to search?
Tell us which systems, archives and exports you have and what you want to do with them. We’ll reply by email or arrange an initial conversation.
Message sent
Thanks for your message. We’ll be in touch as soon as possible.
Frequently asked questions
It builds the foundation for reports and models: bringing in data from different sources, standardising it and putting it on a platform where everyone uses the same, validated version. It then keeps the data-loading processes running.
You can read how we approach this work on our data engineering page.
No. We work from Westerdoksdijk in Amsterdam. We work with organisations in Groningen, Friesland and Drenthe mostly online, and visit to assess what is needed and test our work with users.
None of the case studies here took place in Groningen; they show what we build.
First, we look at what is there: which fields, periods and codes. Then we put the data on the platform unchanged and build a clean version alongside it, using the same definitions as the rest of the data.
That preserves the original and lets you trace where a value came from.
It is an overview of all the tables on the platform, with a description, an owner and access permissions for each one. We used Unity Catalog for Concertgebouworkest. Without a catalogue, only the person who built the platform knows where to find things.
You can read more about ownership and agreed standards on our data governance page.
dbt records transformations as code: from raw source data to clean tables, with tests and documentation. Each step can be reviewed and repeated, and changes follow a defined process.
You can read how we set up dbt on our dbt implementation page.
Often Databricks on Azure, as we are a partner, but we first look at what you already use. Storing data in open formats such as Delta Lake and handling transformations in dbt help you avoid being tied to one provider.
The code stays in your own environment, so another team can take over.
Yes. If you would prefer someone to build alongside your team, read about hiring a data engineer. For a project, we provide the team and deliver the agreed result.
We agree in advance who will manage the platform afterwards.
That is up to you; the platform supports all three. For reporting in the north of the Netherlands, read about our Power BI specialist in Groningen.
For models, there is our data science consultancy in Groningen, and for applications based on language models, our AI consultancy in Groningen.
What do you build on the platform: models or reporting in the North?
- Data engineeringHow we build pipelines and platforms that give reports and models the same data.
- Data science consultancy in GroningenModels that predict or classify for organisations in Groningen and the northern Netherlands.
- Power BI specialist in GroningenOne Power BI model for organisations with locations across the North.