QA review helps us maintain the quality of data solutions in Power BI, Microsoft Fabric and SQL. Thanks to an independent perspective, we can identify technical debt early, improve maintainability and give the customer confidence that the solution will still hold up in a year or two.
At intecs, we have a standard code review process. Developers and consultants continuously review code directly on projects, whether it involves SQL, data models, Power BI reports or solutions built on Microsoft Fabric.
High-quality development of data solutions would not work without code review. At the same time, we know that code review alone is not enough.
When someone is deeply involved in a project for a long time, they gradually lose distance. They know the context, understand historical decisions and know why certain things were done in a specific way. But that is exactly why they can easily overlook problems that are not urgent now, but may start complicating further development over time.
Typically, this means technical debt. It can appear quietly. In SQL code, in the data model, in DAX measures, in documentation or in the setup of data quality checks. At first glance, the solution may work well. Reports load, data refreshes and the customer gets the outputs they need.
But the question is: will such a solution still work well in a year or two? That is why we do QA reviews. For the customer, this means confidence that the data solution they invest in is not built only for fast delivery, but also for long-term maintainability, performance and readiness for future development.
What is a QA review and how does it work?
The quality of a data solution is not a coincidence. How QA reviews work at intecs
QA review is an independent technical assessment of a data solution. It is carried out by an experienced consultant or engineer who is not part of the project team. This allows them to look at the solution from a distance and evaluate not only individual parts, but also the overall technical quality.
👉 This is not about checking every single line of SQL. The goal is not to look for mistakes at any cost. A QA review looks at whether the solution is well designed, readable, maintainable and ready for future changes.
We mainly focus on:
- data model design,
- the quality of SQL code and transformation logic,
- compliance with standards and best practices,
- documentation, code comments and versioning,
- notification setup and data quality checks,
- readability of the solution for someone seeing it for the first time,
- performance, costs and options for future development.
For Power BI and Microsoft Fabric projects, we also use tools such as Tabular Editor, DAX Optimizer and Measure Killer.
These tools help us identify faster, for example:
- performance issues in Power BI reports,
- inefficient DAX measures,
- unused measures or columns,
- an overly complex data model,
- areas where technical debt is accumulating.
QA review helps us answer a simple question: will the data solution hold up in the long term? Not only technically, but also from the perspective of costs, operations, scalability and the work of other people who may join the project in the future.
Why we do QA reviews
There are three main reasons:
We help keep technical debt under control
A data solution can work and the customer can be satisfied with it. But that does not automatically mean it is sustainable in the long term.
Technical debt often does not show immediately. At first, it only slows down small changes. Later, it complicates new requirements, extends development time and increases the costs of operations and further development.
QA review helps identify weak spots early. In practice, this means fewer surprises, less improvisation and better control over how the data solution develops. As a result, the customer gets not only a functional output, but a solution that makes sense to maintain and develop.
We share know-how across teams
QA review is not a one-way inspection. It is a way to share experience between people and projects.
The review is carried out by an experienced consultant who brings a different perspective and broader context to the project. Every project is different, but many situations repeat across data solutions. This makes it possible to transfer good practices between teams.
The discussion around the output is also valuable. The team receives feedback, the reviewer gains a better understanding of the project context, and together they look for a solution that makes sense both technically and from a business perspective.
We train skills we use with customers
The ability to quickly understand someone else’s data solution is essential for our work.
With customers, we often enter existing Power BI reports, data warehouses, SQL databases or Fabric environments. We need to understand how the solution is designed, where its strengths are and where risks may arise.
QA reviews help us develop this skill systematically. So this is not only an internal quality check. It is also practical training for real projects where we need to quickly understand the context, define the problem and suggest specific steps for improvement.
What QA review looks like in practice
The output of a QA review is a structured document that summarizes:
- what works well,
- where we see risks,
- what makes sense to improve,
- how important each point is,
- what next steps we recommend.
The document is followed by a joint meeting with the project team. This is not just about handing over a list of comments. We go through the results together, put them into context and discuss what has the biggest impact on the customer and the further development of the project.
After a few weeks, we return to the outputs. We track what has been improved, what remains open and what real impact the changes have. That is why we do not see QA review as an audit where someone comes in, finds mistakes and leaves. It is a form of collaboration. Always with respect to the specific project, customer, technology and development phase.
🎯 The goal is simple:
to have a data solution that is readable, maintainable, performant and ready for further development
to give the customer confidence that they will not be surprised in a year by problems that are not visible today.
Listen to an episode of our intecs insider podcast, where we discuss what it means to build data solutions that help you sleep well at night — not only because they work, but because they are resilient to change, scalable and transparent.
Who we help become data-driven
