System and method for generating scores for predicting probabilities of task completion

A non-transitory machine-readable medium having instructions stored therein, which when executed by a processor, cause the processor to perform operations, the operations comprising:

• receiving, from a data store, a set of features and a workflow for training a first type of ML model, the workflow specifying a data source, a plurality of stages and associated parameters for training the first type of ML models,
• retrieving, from the data source, training data for the set of features,
• segmenting the training data into different segments,
• training a separate first type of ML model using each of the different segments of the training data in accordance with the workflow, and
• persisting the plurality of first type of trained ML models into the data storage.

A method of generating scores to predict probabilities of task completion, comprising:

• receiving a request for scoring a task, the request including a current status of the task;
• determining one of a plurality of pre-determined segments that the task belongs to;
• retrieving, from a cloud storage service, a corresponding machine learning (ML) model that has been trained using training data for the segment that the task belongs to;
• generating, by the corresponding ML model, a probability score, indicating the likelihood that the task is to be closed;
• identifying one or more factors that account for a predetermined portion of the probability score.