Automatic workflow generation and optimization

1. A device, comprising:

one or more memories; and

one or more processors, communicatively coupled to the one or more memories, configured to:

receive training data identifying a set of workflows;

generate a set of workflow templates based on the training data identifying the set of workflows,

wherein a workflow template, of the set of workflow templates, is associated with a set of steps for completing a task associated with the workflow template;

receive, from a client device associated with an entity, a new task for automation using a new workflow;

parse the new task to identify one or more steps associated with the new task;

process information associated with the one or more steps to identify one or more workflow templates, of the set of workflow templates, satisfying a matching score threshold with the new task;

generate a workflow recommendation relating to the one or more workflow templates; and

output workflow data associated with the workflow recommendation.

2. The device of claim 1 , wherein the one or more processors, to parse the new task to identify the one or more steps, are configured to:

extract task data from the client device;

process the task data to generate a set of work items; and

identify the one or more steps based on the set of work items.

3. The device of claim 1 , wherein the one or more processors are further configured to:

train a workflow generation model using the training data; and

wherein the one or more processors, to parse the new task to identify the one or more steps, are configured to:

parse the new task using the trained workflow generation model.

4. The device of claim 1 , wherein the one or more processors are further configured to:

monitor a data stream of workflow completion data regarding completion of a new workflow associated with the workflow recommendation; and

update the new workflow during completion of the new workflow based on monitoring the data stream.

5. The device of claim 4 , wherein the one or more processors are further configured to:

detect an event associated with the new workflow; and

wherein the one or more processors, to update the new workflow, are configured to:

update the new workflow based on detecting the event.

6. The device of claim 4 , wherein the one or more processors are further configured to:

receive a change request associated with the new workflow; and

wherein the one or more processors, to update the new workflow, are configured to:

update the new workflow based on the change request.

7. A method, comprising:

receiving, by a device, training data identifying a set of workflows;

generating, by the device, a set of workflow templates based on the training data identifying the set of workflows,

wherein a workflow template, of the set of workflow templates, is associated with a set of steps for completing a task associated with the workflow template;

receiving, by the device and from a client device associated with an entity, a new task for automation using a new workflow;

parsing, by the device, the new task to identify one or more steps associated with the new task;

processing, by the device, information associated with the one or more steps to identify one or more workflow templates, of the set of workflow templates, satisfying a matching score threshold with the new task;

generating, by the device and using a recommendation engine, a workflow recommendation relating to the one or more workflow templates; and

outputting, by the device, workflow data associated with the workflow recommendation.

8. The method of claim 7 , further comprising:

receiving a negative response to the workflow recommendation; and

feeding the negative response to the workflow into the recommendation engine to retrain the recommendation engine.

9. The method of claim 8 , further comprising:

generating a new workflow recommendation based on feeding the negative response to the workflow into the recommendation engine to retrain the recommendation engine; and

outputting information identifying the new workflow recommendation.

10. The method of claim 7 , further comprising:

monitoring a data stream to generate a data set regarding completion of a workflow associated with the workflow recommendation; and

feeding the data stream into the recommendation engine to retrain the recommendation engine.

11. The method of claim 10 , further comprising:

generating a new workflow recommendation based on feeding the data stream into the recommendation engine to retrain the recommendation engine; and

outputting information identifying the new workflow recommendation.

12. The method of claim 10 , wherein the data set regarding completion of the workflow includes at least one of:

a path traversed within the workflow,

an amount of time spent on a process step of the workflow,

an idle time during completion of the workflow,

a completion of a process step of the workflow, or

a failure of completion of a process step of the workflow.

13. The method of claim 7 , wherein parsing the new task to identify the one or more steps comprises:

parsing the new task using at least one of:

a regression random forest algorithm,

a support vector regression algorithm, or

a naive Bayes classification algorithm.

14. The method of claim 7 , wherein processing the information associated with the one or more steps to identify the one or more workflow templates comprises:

processing the information associated with the one or more steps using at least one of:

a regression random forest algorithm or

a support vector regression algorithm.

15. The method of claim 7 , further comprising:

optimizing a new workflow associated with the workflow recommendation using at least one of:

a K-means clustering algorithm, or

a discriminant analysis algorithm.

16. The method of claim 7 , wherein the workflow recommendation includes information identifying at least one of:

a step duration for a particular step in a new workflow,

a process path for the new workflow,

a quantity of times that the particular step is performed in the new workflow,

a quantity of active users of the new workflow,

the one or more templates associated with the new workflow, or

an idle time associated with the new workflow.

17. A non-transitory computer-readable medium storing a set of instructions, the set of instructions comprising:

one or more instructions that, when executed by one or more processors of a device, cause the device to:

receive, from a client device associated with an entity, a new task for automation using a new workflow;

parse the new task to identify one or more steps associated with the new task;

identify one or more workflow templates satisfying a threshold matching score with the one or more steps;

create a new workflow based on the one or more workflow templates;

add one or more new steps to the new workflow;

associate one or more assets with a step of the one or more new steps;

generate one or more rules for the new workflow based on the one or more assets;

select one or more trigger points for the new workflow based on the one or more rules; and

set up one or more notifications for the new workflow based on the one or more trigger points.

18. The non-transitory computer-readable medium of claim 17 , wherein the one or more instructions, that cause the device to creating the new workflow, cause the device to:

create the new workflow based on processing data associated with the new task using a plurality of prediction models.

19. The non-transitory computer-readable medium of claim 18 , wherein the plurality of prediction models includes at least one of:

a template predictor,

a workflow predictor,

a step predictor,

an asset predictor,

a rule predictor,

a trigger predictor, or

a notification predictor.

20. The non-transitory computer-readable medium of claim 18 , wherein the plurality of prediction models includes a first model using a first type of algorithm and a second model using a second type of algorithm that is different from the first type of algorithm.