Traditional approaches to RPA have long been used to automate repetitive tasks within organizations. However, they remain limited by their dependence on specialized knowledge and their inability to handle complex decision-making tasks. While LLMs have recently enabled new approaches to automation, solutions that incorporate them often suffer from constraints, such as their domain specificity, their dependence on large proprietary datasets, or the generation of code or programs in natural language rather than fully executable workflows. We present Text2Workflow, a generalized approach that automatically converts natural language queries into complete, executable workflows. Text2Workflow provides a domain-independent framework capable of supporting a wide variety of business processes without requiring fine-tuning to domain-specific datasets.