Methodology for objective comparison of DataOps platforms

This thesis focuses on creating a methodology for objective comparison of DataOps platforms in the context of organizations' growing dependence on efficient data flow management. The current DataOps tools market is characterized by rapid technological development and the absence of a unified framework for systematic evaluation, which complicates organizational decision-making when selecting an appropriate platform. The main objective is to create a flexible and reproducible methodology based on the principle of contextual objectivity, combining expert insights from practice with measurable technical parameters. The thesis employs a mixed-methods research design including a qualitative phase with semi-structured interviews with three industry experts and a quantitative phase with experimental testing using the TPC-H benchmark. The result is an evaluation methodology structured into five main dimensions (technical efficiency, data quality and governance, CI/CD and automation, user experience, business impact) encompassing a total of twenty metrics. The methodology was applied to three representative platforms – Keboola, Microsoft Fabric, and Databricks – and implemented in a publicly available Streamlit application. Validation with experts confirmed the practical applicability of the methodology and identified specific application scenarios in practice. The thesis provides a structured approach to evaluating DataOps platforms that respects the diversity of organizational contexts and enables informed decisions based on the specific priorities of a particular organization.