Rīga Stradiņš University uses Fabric to turn fragmented data into intelligent insights

Rīga Stradiņš University (RSU) is Latvia's leading university fully integrated into the healthcare system, with more than 12,000 students from over 80 countries, over 600 annual scientific publications, and an expected €1 billion contribution to Latvia's economy by 2030, RSU continues to expand its academic and research excellence.

As the university grew rapidly, so did the complexity of its data landscape—creating a new challenge: how to turn an ever-growing volume of data into meaningful insights.

"We always have some growth challenges to solve," explains Zigmunds Zitmanis, Director of IT, describing RSU's unique situation of continuous expansion. 

One result of these "growth challenges" was the emergence of around 50 different systems across multiple domains—including learning management, student information, research, and enterprise resource planning—what Zitmanis calls a "modular IT architecture”. These systems had evolved into a maze of fragmented data ecosystem, making it difficult to achieve a comprehensive view of their data.

To better understand its digital strengths and gaps, RSU collaborated with an external consulting partner, whose digital maturity assessment revealed that the university’s data infrastructure was not keeping pace with its ambitions. These findings highlighted the need for a more unified and forward-looking data strategy.

The human cost of missed data and missed moments

The human cost of this fragmentation was a limited ability to anticipate and improve student retention. “We faced a challenge with student dropouts and were not able to find out when these dropouts were going to happen, or the reasons why,” says Toms Baumanis, the university's chancellor. This wasn't just a statistical concern; it represented real students whose academic careers might have been saved with timely intervention. "Sometimes the cause isn’t a lack of study capacity, but personal difficulties affecting students’ well-being, whether family-related, or health-related,” Baumanis explains. “Those can be solved. But very often, in big organizations, you miss that moment."

Every academic year, RSU collected study evaluation questionnaires from students, containing up to 100,000 open-text comments—a massive volume of qualitative feedback offering deep insight into the student experience. Yet, as Baumanis admits: "No human is able to analyze those." The sheer volume of unstructured data made it difficult to respond to student concerns in a timely manner.

An all-in-one platform for data intelligence

RSU's relationship with Microsoft had been evolving for over a decade. When Microsoft Fabric launched, the university’s IT team immediately recognized its potential to unify data management and analytics within a single, intelligent platform. 

"We finally saw that there is a platform where everything is built-in—a platform which addresses all data management challenges—from metadata and integration management to data transformations," Zitmanis explains. 

RSU was the first educational institution in the Baltics to implement Microsoft Fabric on such a scale, introducing new standards in data integration and management. Gatis Pralics, RSU Information System Architect, explains why Fabric was the right fit for their vision.

"Microsoft Fabric shifts the focus," he says. "It's a cloud-based, all-in-one platform for data engineering and analytics. This unified platform brings together all the parts we need. It's a single, cloud-based environment, so it eliminates the need to use some external tools and install something like servers, switches, operating systems, and so on. We can basically go straight to data engineering and data analytics."

The decision wasn't made in isolation. RSU conducted a Microsoft-funded proof of concept with partners, converting existing Power BI reports to Microsoft Fabric. "It was easy and fast,” Pralics recalls. “We were very surprised. It was a real ‘wow’ moment.” The collaborative effort between Microsoft and its partners significantly accelerated the implementation process. 

Translating student feedback into action

One of RSU's most innovative implementations involved using Azure OpenAI and Microsoft Fabric to improve the analysis of student feedback. Through an iterative development process, Pralics and the team refined the AI system's accuracy by testing it with datasets of hundreds to thousands of student comments, evaluating the results through human review, and adjusting the model through nearly ten cycles of improvements. The sentiment analysis system processes approximately 100,000 open-text comments each academic year, reaching an impressive level of accuracy—closely aligning with human evaluations in approximately 95% of cases, based on internal testing.

The implementation radically improved how RSU visualized course quality, with green flags instantly highlighting positive feedback and red flags alerting administrators to concerns that required attention. This rapid identification capability introduced a new level of transparency and responsiveness in faculty accountability. Detailed instructor responses to student feedback became visible to future students, fostering continuous improvement.

Uncovering hidden issues and measuring success

The power of integrated analytics became evident as RSU uncovered challenges they had not been aware of before. By analyzing student feedback about learning environments, the university identified specific areas where space utilization and accessibility could be improved. This insight allowed RSU to make data-driven adjustments to enhance the overall student experience.

Another revelation came from exam integrity analysis. Student feedback revealed widespread “inconsistencies,” but not due to student dishonesty. Rather, Baumanis discovered, "the pool of multiple-choice questions was too small because the random selection of questions was repeating too often.” These insights led to fundamental improvements in exam design, question pool expansion, and scheduling processes, directly enhancing the student experience.

RSU’s success at better managing and making use of their data can be measured both quantitatively and qualitatively. The plan to have all reports integrated into one system by the end of the calendar year represents a fundamental shift in how RSU typically operates. This integration enhances the university’s ability to make data-driven decisions and improve overall efficiency across all departments.

"We see that problems with data accuracy have decreased,” says Zitmanis, “from things like 5% of all students having some minor data accuracy problems in their portfolio to something like 2%. So, the number of students that get everything properly right and without interruptions has increased, which is important.”

The university is now able to monitor financial metrics at a more granular level, supporting more informed decisions around program viability and staffing. Looking ahead, RSU is already starting to see the benefits of using analytics to identify potential issues earlier and enable timely interventions for at-risk students. “Now we are at the moment where we can really see when the problem is growing," Baumanis states. "These analytics we use for studies are detailed enough to address our needs at operational level, and that is most important."

Reducing bureaucracy and boosting efficiency

Within the next 6-8 months, Zitmanis outlines plans to complete the migration of all reporting to Fabric and implement Microsoft Purview for enhanced data governance. 

"We aim to transform our processes regarding our data-driven decision making,” says Baumanis. “At the moment we have 40 processes—too many. We want to go down. And when you have a digital process...our assumption is then that you will have fewer quality issues, which will bring you less bureaucracy and more efficient administration." 

The university has approved creating two new units—one for IT security and another for artificial intelligence—to support this evolution. Having evolved from a cautious adopter into a regional champion for Microsoft solutions, RSU’s presentation “How to Gather Insights from Data - RSU Microsoft Fabric Use Case” at conference RSU Research Week 2025 sparked significant interest from other universities across the education sector.

For institutions considering a similar transformation, RSU's experience offers valuable insights. Baumanis emphasizes the fundamental importance of data quality: "I think everybody has to pay huge attention to data quality. Data has to be clean or pure because we have spent much time sorting out which data is real and can be integrated with other data." Like many institutions racing to implement new data platforms, RSU faced the fundamental dilemma of balancing immediate business needs with long-term architectural planning.

“It’s a classic chicken-and-egg dilemma,” says Zitmanis. “On one hand, we want to move fast and provide more data to Fabric to generate insights. On the other, we need internal guidelines for how to structure that data to ensure it delivers maximum value.”

As the first education customer in the Baltics to implement Microsoft Fabric at this scale, RSU now champions a bold vision. The university continues advancing its data analytics capabilities, expanding reporting accessibility, introducing organization-wide KPIs, and launching a unified data portal. As Zitmanis reflects, "Microsoft Fabric gives us confidence in our data strategy for the future. It's a secure, feature-rich, and flexible platform that lets us grow without limitations."

Discover more about Rīga Stradiņš University on Facebook and YouTube.