In the rapidly evolving fields of computer education and information technology, it is vital to effectively educate future engineers in critical skills such as programming, system development, and project management principles. This study, conducted at Riga Technical University (RTU), integrates innovative teaching methods, including a “Learning Style” test, to improve group project outcomes in several courses. The analysis spans courses like “Algorithmization and Programming of Solutions” and “Development of Web-Applications for the Internet,” assessing the impact of team composition based on learning styles and the optimal group sizes for collaborative projects. Results from the 2024 academic year indicate that larger groups (4-7 members) tend to achieve higher project grades compared to smaller ones. Interestingly, while no significant difference in grades was observed across different learning styles, the project grades varied with the number of students per group within certain learning styles, particularly among ‘Practitioners’ and ‘Logicians’. This study highlights the potential benefits of structured group formation and suggests further research into the effects of group composition on project success, proposing periodic progress reports as a novel approach for future courses. Thus, it contributes significant insights relevant both to pedagogical methods and management in educational settings.