Article
For Perttu Salonen, the practicality behind the title of Senior Data Architect depends on the day, the client and one’s appetite for taking on new challenges. The everyday work of a Nitorean data professional is not static repetition, but rather AI era problem-solving that evolves with the client’s goals.
“The work varies tremendously. There is no premade template for a data architect’s day. My tasks often revolve around identifying the client’s most substantial challenges and desired outcomes and uncovering the best tools and methods to tackle each one. The hands-on aspect changes as the client’s wishes, goals and priorities change. I’ve been with the same client for the past four years, but the work itself has changed shape many times over with new challenges, objectives and constantly evolving tools,” Salonen begins.
Salonen works primarily with data platforms, such as Databricks, GCP/BigQuery, dbt, and Airflow, and is used to jumping into a project in an advisory role. However, he describes himself as a hands-on architect first, eager to tackle practical challenges.
“I don’t just want to draw up plans and run slideshow sessions. I want to get my hands dirty and solve real-world problems. In other words, develop, implement and code.”
A data professional must strive for adaptability
According to Salonen, one golden rule of development keeps a data architect’s work particularly varied: the further you get, the further you want to go. Plans often change as goals expand and become more defined. This frequently acts as a catalyst for both expanding original development goals and launching entirely new initiatives. This means a data professional must be able to shift focus.
As an example, Salonen mentions designing a data lineage solution built on top of a client’s data platform, enabling the organisation to identify where data originates and trace its paths. The groundwork took a long time, as the design phase involved interviews, extensive background research and an assessment of potentially suitable tools. Then the client began a major cloud migration. Services that had been running on Kubernetes in their own data centres moved to GCP, and many things had to be redesigned again.
“This work requires adaptability. You can’t let yourself get stuck patching over the same stretch of real estate for comfort’s sake. You always have to be ready to respond to the client’s changing needs and apply your expertise in new ways,” Salonen sums up.
The unique perspective of a consulting role
Salonen notes that a significant part of a data architect’s role is filtering information and making sure the necessary knowledge reaches the right teams and people within the client organisation. Tacit, fragmented and poorly conveyed information creates sturdy invisible barriers that make decision-making harder and hold projects back.
“A consulting role allows you to complement the organisation’s internal perspectives and act as a kind of communication glue between different teams and stakeholders. Areas of improvement and bottlenecks are often easier to spot when you can look at them through the lens of an outsider,” Salonen explains.
A key part of the day-to-day of a consulting data architect is to internalise the client organisation’s culture, ways of working and team dynamics, and ensuring that the right information reaches the right people. The numerous variables in the equation range from the company’s area of industry to its age and size. At times, identifying pressure points calls for a great deal of initiative.
“Early in my Nitor career, I spent some time working with a gaming company where the environment called for a much more proactive approach than I was used to. I couldn’t just sit back and assume that someone would come tugging at my sleeve with a ready-made list of concerns and needs. The solution was simple: get to the office more often. Meet people and engage in conversation. It was an important lesson in how essential active communication is for success in this job,” Salonen emphasises.
The role of data has changed enormously
According to Salonen, attitudes towards data have changed significantly in recent years. The focus has shifted from solving challenges with data storage to exploring the best ways clients can utilise their data and finding the most valuable use cases.
“Companies have come to grips with the fact that data needs to be of high quality before you can reliably act on it. Tools have also evolved tremendously over the years: you can get a capable data platform up and running without a massive financial investment. Cloud migration is also accelerating development, as it is no longer necessary to invest in in-house hardware. The platform can be scaled in direct proportion to how the business develops over time,” Salonen notes.
AI agents and tools are also rapidly becoming more widespread and versatile, affecting the daily life of data architects. Salonen believes that many clear and simple tickets can be resolved with automation in the future, with AI handling some of the straightforward tasks autonomously and humans focusing on reviewing the quality of the end result.
“The biggest challenge for AI agents is their poor grasp of context. For example, AI can’t read the wishes and intentions between the lines of a brief the way a human can. Particularly in large, complex systems, human expertise is becoming even more critical, especially in the design phase. Once the plans have been carefully worked out and tested in practice, bigger pieces of the whole can be handed over to AI and automation,” Salonen explains.
The increasing role of AI agents on the data architect front
When asked about the future of the role of data architect, Salonen calls it the million-dollar question with a laugh. When it comes to the direction of his own professional development, however, he has a very clear vision.
“My own path of development is leading towards a more holistic use of agents. I’m particularly interested in how we can automate and build processes where knowledge and need find each other through less discord. Especially in long-established companies tacit knowledge may have accumulated over decades, which can make leveraging data really, really challenging at times,” Salonen elaborates.
Salonen notes that focusing on automation and seamless data management helps clients free up more resources for work that cannot be automated. In the future, there will also be a growing need for insight into how agentic solutions are built and what fundamental prerequisites the work involves.
“Technological development is only accelerating, and data architects will continue to have more work than there are hours in the day. Still, the most important thing about this job is to truly care about what you’re doing. That means sometimes questioning things and being ready to propose a completely different approach or implementation from the one you originally set out with,” Salonen describes.
At the end of the interview, Perttu returns to what keeps him motivated as a data architect day after day.
“I feel I’m in a privileged position at Nitor. Day after day, I get to focus on solving genuinely interesting and motivating challenges and work with highly skilled people on fascinating client engagements with modern tools. The great employee benefits are a nice bonus, too,” Salonen says with a laugh.
