Reproducibility is a fundamental requirement in scientific computing, ensuring that computational and experimental results can be replicated independent of time, system, and user. This thesis compares two reproducibility platforms - Binder, and CodeOcean - based on usability, performance, limitations, and suitability for different application areas. The study focuses on three domains: Business Analytics, Machine Learning and Scientific Computing.
To systematically assess these platforms, the thesis will implement standardized "Hello World" test programs tailored to each domain. These programs will be deployed and executed on all platforms to analyze reproducibility challenges, including dependency management, execution consistency, and ease of use. The final results will provide insights into the strengths and weaknesses of different reproducibility solutions and offer recommendations on selecting the right platform based on specific research and industry needs.