creating a challenges-successes-failures database storing cleaned and processed data collected from a history of well completions for wells, wherein the challenges-successes-failures database identifies: challenges encountered during the well completions, successes and failures that resulted during the well completions corresponding to the challenges, and job parameters used during the well completions;

selecting a dimension embedding algorithm to be used to represent the cleaned and processed data;

performing hyper-parameter tuning on the dimension embedding algorithm;

generating, using the dimension embedding algorithm, a dimension embedding model;

adding the dimension embedding model to a system pipeline for a new well completion job;

receiving new job parameters corresponding to the new well completion job;

running, using the challenges-successes-failures database and the new job parameters, nonlinear dimension embedding algorithms against data points in the cleaned and processed data collected from the history of well completions;

generating, in response to running the nonlinear dimension embedding algorithms, scatter plots of two-dimensional (2D) points labeled with each point's job parameters for the new well completion job and the wells in the past; and

executing a well completion operation for the new well completion job using a labeled job parameter of the scatter plots of 2D points.