Dynamic analytics and forecasting for messaging staff

A system comprising:

• memory that stores historical messaging data;
• a graphical user interface that presents a plurality of skills and a plurality of target service levels;
• and
• a processor that executes instructions stored in the memory, wherein execution of the instructions by the processor:
• receives a selection via the graphical user interface, wherein the selection is of a skill and a desired target service level for the skill corresponding to a desired customer satisfaction level;
• filters the historical messaging data based on the selected skill and the selected target service level for the skill to obtain filtered results, wherein the filtered results include a set of the historical data associated with the selected skill and the selected target service level for the skill;
• analyzes, using the filtered results, a current messaging session in real-time as messages of the current messaging session are being routed, wherein the current messaging session is analyzed according to one or more skills being exercised during different time periods within the current messaging session;
• uses the analysis to attribute a particular time period to the selected skill in real-time, the particular time period corresponding to a time period during which the selected skill is being exercised within the current messaging session;
• associates the particular time period during which the selected skill is being exercised with a particular customer satisfaction level;
• generates a prediction by applying a trained learning model to the filtered results, the particular time period attributed to the selected skill, and the associated particular customer satisfaction level, wherein the prediction corresponds to a projected staffing level for servicing messaging conversations associated with the selected skill and the target service level for the skill, and wherein the learning model is refined according to the selected skill associated with the particular time period;
• and
• dynamically adjusts a staffing forecast in real-time based on the prediction, wherein the staffing forecast is used to determine staffing related to the skill needed to achieve the target service level for the skill.

A non-transitory, computer-readable storage medium, having embodied thereon a program executable by a processor to perform a method comprising:

• presenting a plurality of skills and a plurality of target service levels for skills via a graphical user interface;
• receiving a selection via the graphical user interface, wherein the selection is of a skill and a desired target service level for the skill corresponding to a desired customer satisfaction level;
• filtering historical messaging data based on the selected skill and the selected target service level for the skill to obtain filtered results, wherein the filtered results include a set of the historical data associated with the selected skill and the selected target service level for the skill;
• analyzing, using the filtered results, a current messaging session in real-time as messages of the current messaging session are being routed, wherein the current messaging session is analyzed according to one or more skills being exercised during different time periods within the current messaging session;
• using the analysis to attribute a particular time period to the selected skill in real-time, the particular time period corresponding to a time period during which the selected skill is being exercised within the current messaging session;
• associating the particular time period during which the selected skill is being exercised with a particular customer satisfaction level;
• generating a prediction by applying a trained learning model to the filtered results, the particular time period attributed to the selected skill, and the associated particular customer satisfaction level, wherein the prediction corresponds to a projected staffing level for servicing messaging conversations associated with the selected skill and the target service level for the skill, and wherein the learning model is refined according to the selected skill associated with the particular time period;
• and
• dynamically adjusting a staffing forecast in real-time based on the prediction, wherein the staffing forecast is used to determine staffing related to the selected skill needed to achieve the target service level for the skill.