The rapid advancement of Artificial Intelligence (AI) necessitates preparing the next generation to be AI- and data-literate citizens and problem solvers. Despite efforts to integrate AI into K-12 education, many teachers lack the expertise to create meaningful AI learning experiences. We developed ML4Inq, an innovative online authoring tool that enables secondary school teachers to create data-driven inquiry-based learning (IBL) activities with machine learning (ML) technologies such as clustering and classification. Findings of a co-design workshop with 14 secondary school teachers show that teachers successfully created a diverse range of curriculum-aligned activities using ML4Inq, in collaboration with ML experts. We discovered key connections between ML-revealed patterns and IBL behaviors, offering design implications for future tools that support ML practices in secondary schools.