Workflow insight engine and method
A computer-implemented method is disclosed for generating insights for improving software workflows, where a workflow corresponds to a sequence of interactions of a user with one or more different user interface screens of software applications to perform a task. Attributes of the workflow associated with quality, efficiency and other attributes are measured by scoring aspects of the workflow and generating reports. The reports may also provide insights on opportunities to automate workflows.
1. A computer-implemented method comprising:
monitoring, via an insight engine, instances of a workflow associated with user-interactions with two or more different user interfaces of two more different software applications used in a sequence of keystrokes and mouse clicks navigating a plurality of screens to perform a work task having a sequence of steps;
scoring, via the insight engine, the workflow based on at least one scoring rule to evaluate the workflow;
generating insights on workflow patterns identified by the scoring to identify opportunities to automate the workflow; and
utilizing a machine learning model to automate the workflow based on the identified opportunities.
2. The computer-implemented method of claim 1 , wherein the scoring comprises generating a metric for the workflow associated with at least one of: 1) an efficiency metric of the workflow, 2) a waste metric of the workflow associated with a user interrupting a workflow with non-work related activities; 3) a complexity metric of the workflow, and 4) a friction metric indicative of a level of user effort associated with keystrokes and mouse clicks required by the workflow.
3. The computer-implemented method of claim 1 , wherein the scoring comprises scoring the workflow based on a determination that the workflow uses a software application associated with a risk of user error or unnecessary user steps.
4. The computer-implemented method of claim 1 , wherein the scoring comprises scoring the workflow based on a user using an unstructured application for at least one step of the workflow.
5. A computer-implemented method comprising:
monitoring, via an insight engine, instances of a workflow associated with user-interactions with two or more different user interfaces of two or more different software applications in a sequence of keystrokes and mouse clicks navigating a plurality of screens to perform a work task having a sequence of steps;
scoring, via the insight engine, the workflow based on at least one scoring rule to evaluate the workflow, wherein the scoring comprises generating a workflow path variability score based on a variability in different paths taken in different instances of the workflow;
generating insights on workflow patterns identified by the scoring;
in response to the workflow path diversity exceeding a threshold, utilizing a machine learning model to automate the workflow.
6. A computer-implemented method comprising:
monitoring, via an insight engine, instances of a workflow associated with user-interactions with two or more different user interfaces of two or more different software applications used in a sequence of keystrokes and mouse clicks navigating a plurality of screens to perform a work task having a sequence of steps;
scoring, via the insight engine, the workflow based on at least one scoring rule to evaluate the workflow, wherein the scoring comprises generating a data movement score associated with copy and paste user interface operations of the workflow;
generating insights on workflow patterns identified by the scoring to identify opportunities to automate the workflow; and
utilizing a machine learning model to automate the workflow based on the identified opportunities.
7. The computer-implemented method of claim 1 , wherein the scoring comprises generating an efficiency score to reflect the fraction of time in a workflow in which productive work is performed.
8. The computer implemented method of claim 1 , wherein the scoring comprises generating a complexity score based in part on a number of unique user interface screens in a workflow.
9. A computer-implemented method comprising:
monitoring, via an insight engine, instances of a workflow associated with user-interactions with two or more different user interfaces of two or more different software applications used in a sequence of keystrokes and mouse clicks navigating a plurality of screens to perform a work task having a sequence of steps;
scoring, via the insight engine, the workflow based on at least one scoring rule to evaluate the workflow, wherein the scoring comprises generating a friction score indicative of a burden on user to navigate and enter data and commands in user interfaces through keystrokes and mouse clicks;
generating insights on workflow patterns identified by the scoring to identify opportunities to automate the workflow; and
utilizing a machine learning model to automate the workflow based on the identified opportunities.
10. A computer-implemented method, comprising:
monitoring, via an insight engine, instances of a workflow associated with user-interactions in a sequence of keystrokes and mouse clicks with two or more different user interfaces of two or more different software applications used to complete a work task having a sequence of steps;
scoring, via the insight engine, the workflow based on at least one scoring rule to evaluate the workflow;
generating, via the insight engine, an insight about at least one problem with at least one workflow pattern identified by the scoring to identify opportunities to automate the workflow;
reporting on the identified opportunities, based on the insight, to automate at least a portion of the workflow associated with the at least one problem; and
utilizing a machine learning model to automate the workflow based on the identified opportunities.
11. The computer implemented method of claim 10 , wherein the scoring comprises at least one of: 1) an efficiency metric of the workflow, 2) a waste metric of the workflow associated with a user interrupting a workflow with non-work related activities; 3) a complexity metric of the workflow, and 4) a friction metric indicative of a level of user effort associated with keystrokes and mouse clicks required by the workflow.
12. The computer implemented method of claim 10 , wherein the workflow comprises a sequence of steps performed using at least two different software applications.
13. A computer-implemented method, comprising:
monitoring, via an insight engine, instances of a workflow associated with user-interactions in a sequence of keystrokes and mouse clicks with two or more different user interfaces of two or more different software applications used to complete a work task having a sequence of steps;
scoring, via the insight engine, the workflow based on at least one scoring rule to evaluate the workflow, the scoring further comprising generating an algorithm for the scoring based on analyzing workflow instances, identifying statistical metrics associated with the analyzed workflow instances, and generating a heuristic scoring rule based at least in part on the statistical metrics;
generating, via the insight engine, an insight about at least one problem with at least one workflow pattern identified by the scoring to identify opportunities to automate the workflow; and
utilizing a machine learning model to automate the workflow based on the identified opportunities.