This study presents the first steps for creation of ACCRa (Auto-Configuration of the Constructor Research Platform), an AI system that intends to automate research platform adoption and configuration. ACCRa is utilized for two key purposes: extraction and generation of code from research papers, and automating configuration of the Constructor Research Platform from source code existing in repositories or described in research papers. Using natural language processing, repository analysis, and dynamic optimization of resources, ACCRa decreases the amount of manual intervention from researchers, thereby addressing a central reproducibility issue of computational research environments. The ability of the system to dynamically adjust to fluctuating resource demands and runtime environments enables increased research productivity and reproducibility of experiments.