Learning based metric determination for service sessions

A non-transitory computer storage medium encoded with computer program instructions that when executed by one or more computers cause the one or more computers to perform operations comprising:

• training a neural network to process a first record of communications between a first service representative (SR) and a first individual during a first service session to generate one or more predicted survey scores that rate the first service session on one or more criteria, wherein the training is conducted using training data comprising, for each of one or more previous service sessions, a respective previous record of the previous service session and a respective set of one or more actual survey scores provided by a serviced individual to rate the previous service session on the one or more criteria;
• receiving, by at least one processor, a second record of communications between a second service representative (SR) and a second individual during a second service session;
• processing, using the trained neural network, the second record to determine one or more new predicted survey scores that rate the second service session, wherein the trained neural network comprises a language neural network model and an acoustic neural network model, and wherein the trained neural network comprises different neurons, each neuron corresponding to a respective criterion of the one or more criteria;
• associating, by at least one processor, the one or more new predicted survey scores with the second individual;
• and
• communicating, by at least one processor, the one or more new predicted survey scores for presentation through a user interface of a computing device.

A system comprising:

• one or more computers and one or more storage devices storing instructions that are operable, when executed by the one or more computers, to cause the one or more computers to perform operations comprising:
• training a neural network to process a first record of communications between a first service representative (SR) and a first individual during a first service session to generate one or more predicted survey scores that rate the first service session on one or more criteria, wherein the training is conducted using training data comprising, for each of one or more previous service sessions, a respective previous record of the previous service session and a respective set of one or more actual survey scores provided by a serviced individual to rate the previous service session on the one or more criteria;
• receiving, by at least one processor, a second record of communications between a second service representative (SR) and a second individual during a second service session;
• processing, using the trained neural network, the second record to determine one or more new predicted survey scores that rate the second service session, wherein the trained neural network comprises a language neural network model and an acoustic neural network model, and wherein the trained neural network comprises different neurons, each neuron corresponding to a respective criterion of the one or more criteria;
• associating, by at least one processor, the one or more new predicted survey scores with the second individual;
• and
• communicating, by at least one processor, the one or more new predicted survey scores for presentation through a user interface of a computing device.