NIRAS A/S
NIRAS A/S is an international consultancy firm providing engineering, planning and project management services for sustainable development and growth. NIRAS A/S works with clients in various sectors including infrastructure, energy, environment, and food and agriculture. NIRAS A/S is committed to being an attractive high-tech workplace delivering innovative and quality solutions to its customers.
Technological front runner
NIRAS A/S wants to be a frontrunner also in the use of digital technology. Therefore, NIRAS A/S also been keen to explore the possibilities of strengthening their data capabilities and insights with Microsoft Fabric, which is a relatively new platform.
That's why Project Director Sarah Arndt Christensen and two of her colleagues from NIRAS A/S participated in Fellowmind's Fabric Labs at Microsoft, learning about the features and benefits of Microsoft Fabric. They quickly realised how Fabric could help NIRAS A/S achieve their goals.
"Fabric Labs worked really well as a day of learning where my colleagues and I could spar with other developers and try out all the different features and functions on our own data. At the same time, we got some good advice on how to use Fabric in our daily work," says Sarah Arndt Christensen.
A good start to the Fabric journey
After attending Fabric Labs, NIRAS A/S decided to start their Fabric journey together with Fellowmind, who have helped with the implementation and development of the platform. Fellowmind assisted NIRAS A/S in loading various data sources into Fabric and implementing best practices in data management and data platform design.
After attending the Fabric Labs conference, I partnered with Fellowmind as a knowledgeable mentor to help us embark on our journey with Fabric. This collaboration remains essential in developing our governance strategies and strengthening our Fabric capabilities, which will enable our continued success with the Data and Analytics platform in NIRAS A/S.
In addition, Fellowmind has helped NIRAS A/S with the capacity model and its built-in reporting in Fabric, enabling them to track their consumption and costs effectively.
Fellowmind has an in-depth knowledge of Fabric and its potential, and understands how it can be customised to meet a range of business needs, strengthening our BI team at NIRAS A/S. Their close partnership with Microsoft allows them to utilise valuable resources and support to amplify Fabric initiatives.
New data exploration options
Fabric has enabled NIRAS A/S to explore their business data in completely new ways. First and foremost, the access to data across different sources and systems is easier, giving better opportunities to compare and analyse data. Fabric enables NIRAS A/S to make better predictions about future events - and data-supported decisions in general.
Fabric has also helped NIRAS A/S through easier role management, ensuring better access control and data security in accordance with the requirements of NIRAS A/S.
The Fabric project included a classic Extract-Transform-Load (ETL) process consisting of three main steps:
- Extract: Retrieve data from source systems
- Transform: Clean, format and structure data according to the desired model
- Load: Store data in the target system or database
The first step was deciding where to get the source data for the semantic model. Fellowmind and NIRAS A/S decided that CatalystOne, a cloud-based HR system, would be the primary data source, and that it would serve as the "point of truth" or starting point for the entire HR dashboard.
The challenge was that data had to be pulled through an API call, generating one JSON file per data field. The data was then collated into two RAW areas for type 1 and type 2 data, respectively. Type 1 data always overwrites existing data, while type 2 data stores historic information such as changes in department or salary.
The task has been solved using Fabric and the underlying services, Data-pipeline and Notebooks. Data-pipeline has been used for calling and retrieving data from the API, while Notebooks has been used for decoding the JSON files and selecting the desired objects/data fields for loading into the data model. Data from the API calls is eventually stored in a table in a so-called lake house, a hybrid data storage solution combining the best of data storage and databases.
With the extraction in place, Fellowmind and NIRAS A/S moved on to the next step in the ETL process, transforming and cleaning the data. This involved the following tasks:
- Removing duplicates and errors
- Converting data types and formats
- Applying business rules and logic
- Creating dimensions and facts
- Generating surrogate and foreign keys
These tasks were performed using Fabric Notebooks, allowing Fellowmind and NIRAS A/S to write SQL and Python code for manipulating data and creating the desired data structure.
The final step was to load the transformed data into a Power BI dashboard. This dashboard would enable NIRAS A/S to access and analyse HR data and monitor employee performance and trends. This task was also facilitated by Fabric ensuring seamless integration with Power BI and other Microsoft tools.
Fabric also helped ensure that data was reliable and up-to-date in real-time. By using Fabric, Fellowmind and NIRAS A/S have been able to build a data model for the HR dashboard that fulfils their requirements and expectations and delivers value and insights.
Fellowmind assisted NIRAS A/S in loading various data sources into Fabric and implementing best practices in data management and data platform design.
