KEHA Centre

KEHA Centre’s Data Transformation: Building the Mylly Data Platform

Want to see how Agile Data Engine powers large-scale data transformation? Discover how KEHA Centre built a next-generation data platform to tackle public sector challenges.

KEHA Data

At the heart of Finland’s public sector transformation, the KEHA Centre faced a pivotal moment when employment services underwent a significant reform. As a national government agency, KEHA's responsibilities expanded beyond its traditional role – now overseeing employment services, job centers, and other functions affiliated with The Ministry of Employment and the Economy across Finland’s municipalities. These changes pushed it to reimagine how it handled data and delivered services to its stakeholders. 

The Development and Administrative Services Centre for Economic Development, Transport and the Environment (KEHA-keskus in Finnish or UF-centret in Swedish) operates nationwide with a distributed network model, supported by approximately 800 experts serving clients across Finland. 

The KEHA Centre's services are wide-ranging, from development and administrative services to knowledge management, digitalization efforts, and employment promotion. These services cover 15 EVK Centres, employment regions, municipalities, central agencies, public entities, associations, companies, farms, private customers, and welfare areas.  

People working at computer

Key takeaways

  • KEHA Centre built the Mylly Data Platform on Agile Data Engine and Data Vault to serve municipalities, EVK Centres and central government.


  • ADE brought standardized practices: new team members were operational in three days, and data requests that took weeks are now completed in days.

  • In the latest chapter, Agile Data Agent (ADA) cut Data Vault work from around ten days to one or two, with a person reviewing every output.

  • Read how ADA is used in Mylly

Data transformation for a rapidly evolving landscape

With the increasingly complex operational landscape, simple data management was no longer enough. The KEHA Centre needed seamless data flow across government departments, municipalities, and employment services.

Because their existing systems struggled to keep pace with this new reality, in their data transformation the KEHA Centre team needed to consider:

  • Complex and dynamic data needs: Integrating diverse data sources and enabling real-time analytics proved daunting.
  • Reducing person dependency and inefficiencies: Key personnel were critical for operations, leading to bottlenecks and vulnerabilities.
  • Eliminating fragmented tools and processes: A lack of unified systems hindered agility and adaptability. 

To address these challenges, the KEHA Centre needed a comprehensive, scalable data platform – one that could support the diverse needs of multiple public sector entities. The project required both speed and long-term sustainability – a tall order when you need to build a new platforming solution from the ground up.  

Introducing Mylly: A data platform for the future

The KEHA Centre decided to call this new platform Mylly – which is the Finnish word for “mill.” Given the complexity of the data landscape, the team selected Data Vault as the modeling method to build Mylly Data Platform (Tietoalusta Mylly in Finnish).

However, implementing Data Vault from the ground up under a tight schedule required efficient tooling. As Mylly’s Product Manager, Lea Pitkänen, explains:

"We lived in a much smaller world before. Now, the scenarios are broader and more complex, with multiple systems, integrations, and data transfer methods."

Agile Data Engine: The Game-Changer

In order to implement Data Vault efficiently – without the repetitive manual effort typically required for a robust data model – the KEHA Centre adopted Agile Data Engine (ADE). They chose Agile Data Engine because of ADE’s ability to handle the intricate demands of Data Vault 2.0 modeling as well as providing other key benefits integral to the Mylly Data Platform’s operations. Because of ADE, numerous tasks are handled in the database:

"We handle everything within the database with Agile Data Engine — building, moving data, loading, and managing publication areas. Since Agile Data Engine handles these tasks and more, it has become our key tool for ensuring structure and version control", says Data Warehouse Architect Teemu Väisänen.

productui-perspective-data-lineage@2x

Key ADE features driving Mylly’s success

  • Unified DataOps platform: Consolidates data modeling, orchestration, and monitoring.
  • Version control: Ensures consistency and smooth transitions between testing and production.
  • Automation: Reduces manual oversight, allowing developers to focus on value-adding tasks.
  • Cloning feature: Simplifies complex projects, drastically reducing time and effort.
The cloning feature was a revelation. What would have taken days now takes only a few hours.

Vesa Takala

Data Analyst

A powerful project kickoff

The decision to use Agile Data Engine had a transformative impact from the very start. The team found inspiration in the depth and quality of  ADE Basic Training program.

"Agile Data Engine for Data Engineers was very good. The material was prepared well, and the joint sessions and dailies were a good use of time and fostered a sense of togetherness", says Vesa Takala.

"During the training, I learned to do Data Vault 2.0 modeling – the training was an excellent introduction to both ADE and the Data Vault 2.0 method", adds IT Specialist Kimi Öling.

Meanwhile, Data Warehouse Architect Teemu Väisänen appreciated that this training helps onboard new members. Väisänen also emphasized the value of Agile Data Engine’s advanced course which is great for any team member who might need additional training:

"When onboarding a new team member, it’s a big advantage that Agile Data Engine is the one clear place where all the data work is done. So that’s quite straightforward, and when additional training is needed, Agile Data Engine offers a more advanced crash course that's very effective."

Implementing ADE brings immediate operational clarity

Following the succesful kickoff and training, Agile Data Engine’s implementation immediately clarified and streamlined operations. The KEHA Centre was able to centralize data, streamline processes, and provide actionable insights for agencies like EVK, TE-Services, and the Ministry of Employment and the Economy.

Where previously developers juggled multiple tools and systems, they now had a single, unified platform for all data work. As Toni Mäkelä, Senior Technical Architect for the Mylly Data Platform at the KEHA Centre reflected:

"ADE handles practically everything – from data loads to monitoring and modeling. Without it, we’d need two or three different tools to achieve the same results."

Agile Data Engine delivers results

According to Toni Mäkelä, Senior Technical Architect, ADE introduced standardized practices that ensured consistency, even when external consultants were brought in. These changes are particularly noticeable in how new team members adapt. The learning curve, which is often a significant hurdle in complex data environments, became manageable so new developers can quickly grasp the unique aspects of the system, a skill reinforced by ADE’s training. As a result, both new hires and external consultants could contribute meaningfully within days rather than weeks. 

Key benefits of ADE implementation

  • Faster Onboarding: New team members were operational in three days, thanks to ADE’s intuitive processes and effective training.
  • Increased Agility: Data requests that previously took weeks are now completed in days, enhancing responsiveness to stakeholders.
  • Improved Data Trust: Automation and visibility fostered confidence in data outputs across agencies.
iStock-1406705759

Perhaps most significantly, the KEHA Centre achieved something rare in large-scale data operations - consistency. Whether work was done by internal staff or external consultants, the output maintained a uniform standard. This standardization accelerated their ability to deliver solutions to business stakeholders, meeting their expectations for quick, reliable results. 

 
ADE introduced common practices, ensuring consistency whether work is done by our team or consultants.

Toni Mäkelä

Senior Technical Architect

Mylly Data Platform: The backbone of KEHA’s data ecosystem

The Mylly Data Platform emerged as another crucial piece of the puzzle, serving as the backbone for reporting and monitoring. Together with ADE, KEHA created a robust ecosystem that could handle both routine tasks and complex, custom requirements – all while ensuring compliance with legal and data governance mandates essential to public sector organizations.

One pivotal use case for the KEHA Centre team involved preparing municipalities for new employment services obligations in 2024. Agile Data Engine enabled the Mylly Platform to provide: 

  • Secure data storage and separation: Ensures compliance with precise legal requirements.
  • Efficient migration: Transferring large datasets smoothly between environments.
With ADE’s version control, we securely delivered tailored data packages to municipalities without disrupting existing workflows.

Vesa Takala

Data Analyst

The KEHA Centre’s data operations stand as a testament to successful digital transformation in the public sector. Their environment, though more complex than ever, runs smoothly with reliable data pipelines and seamless integration. What began as a response to organizational change  has evolved into a model for efficient, adaptable data management, ready to meet future challenges in Finland’s employment services landscape.

 

Next chapter: agentic development with Agile Data Agent

Mylly was already mature when Solita consultants working in KEHA Centre's environment started piloting the Agile Data Agent (ADA). Routine work such as staging, standard load patterns and ADE package handling was already standardized. Data Vault modeling stayed manual and expensive: hubs, links, satellites, and the business rules on top of them. Generic AI tools were too general to trust with ADE packages.

ADA operates on ADE packages directly and generates metadata. Because Mylly already runs on standardized practices, that output lands in an established, governed workflow. KEHA brought ADA in under the same approval model it had used for other AI tools, since ADA does not access data differently than tools KEHA had already cleared.

"Going from ten days of work to two days is a massive change," says Teemu Väisänen, Data Warehouse Architect at KEHA Centre. This was the finding of how significant the shift was when ADA was able to take care of the Data Vault grunt work - proving the incredible potential of the approach.  

Key benefits of ADA in Mylly:

  • Faster Data Vault work: Work that previously took around ten days came down to one or two.

  • A stronger starting point: Developers start from a generated foundation and move directly into review and refinement.

  • Consistent results across experience levels: Custom skills steer ADA toward the team's conventions, and junior developers benefit in particular.

A person still reads the output before it goes anywhere. "A human still has to be in between and read the end product before pushing it forward," Väisänen says. The next step is giving the agent more context about what tables and views contain and which rules apply.

Read the full story in the blog post: Running ADA where the governance bar is genuinely high. 

Lessons for Public Sector Data Teams

KEHA’s journey serves as blueprint for public sector organizations looking to modernize their data operations:

  • Unified Tools: ADE consolidates multiple functions, eliminating the need for disparate tools.
  • Agility and Resilience: Supports complex environments while reducing person dependency.
  • Empowered Teams: Simplifies workflows and fosters a collaborative culture
  • Governed AI adoption: Start with a small, well-scoped use case and keep a human reviewing the output.

The KEHA Centre’s successful data transformation resulted in a Finnish trademark for the Mylly Data Platform in 2024. As Lea Pitkänen remarked, “Our work is being benchmarked across the field. Interest in Mylly has grown significantly."

The KEHA Centre's journey with Mylly demonstrates how innovation, powered by the right tools, can transform public sector data management into an impact driver that helps public sector organizations become more data driven and efficient. Today  the Mylly Data Platform, powered by Agile Data Engine, is a cornerstone of the KEHA Centre’s strategy of delivering standardized data products and self-service analytics. This empowers stakeholders to manage operations more effectively ensuring long-term success in an ever-evolving landscape. Agile Data Agent now adds a faster way to build Data Vault on top of that foundation.


ADA is in Private Preview and activated per tenant on request. Contact us to try it on a scoped piece of your own model.

Ready to power up your business with Agile Data Engine?

Contact us