Patent US 11,386,368

Techniques are provided for the optimal matching of teachers to students using machine learning models and analysis engines (MLMAEs). In an embodiment, historical and biographical teacher and student data is collected through a data ingestion layer. As the data is ingested, a process is spawned that automatically builds MLMAEs using a plurality of algorithms. The MLMAEs are automatically tested for accuracy, and the MLMAE with the highest level of accuracy is promoted as the winner. Students enter information through a user interface to request optimal teachers for a subject area, and this information is run against the winning MLMAE which returns a list of optimal teachers. Also provided are techniques for teachers to enter their availability, techniques for MLMAE creators to seamlessly test prospective new MLMAEs, and techniques for teachers to understand why they are not matches for specific segments of students and to receive associated professional development resources.