System and method for predictive technology incident reduction

1. A method for predictive technology incident reduction, comprising:

in an information processing apparatus comprising at least one computer processor:

receiving a change record for a proposed change to a computer application or a computer network infrastructure;

analyzing the proposed change for an adverse potential cross impact on another computer application or a computer system;

predicting a probability of failure and an impact of the adverse potential cross impact on the other computer application or computer system using a model;

in response to a low predicted probability of failure, or a high predicted probability of failure with a low predicted impact:

approving the proposed change; and

implementing the proposed change; and

in response to a high predicted probability of failure and a high predicted impact, rejecting the proposed change.

2. The method of claim 1 , wherein the model is selected by a learning engine based on a prior implemented change and a failure and an impact of the prior implemented change.

3. The method of claim 1 , wherein the model comprises at least one of a Nave Bayes algorithm, a k-nearest neighbors algorithm, logistic regression, and an ensemble classifier.

4. The method of claim 1 , wherein the model is selected by providing test data to a plurality of models, and selecting the model based on results of the test data.

5. The method of claim 1 , wherein the predicted probability of failure and the predicted impact of the proposed change are based on at last one of current change data, change history data, configuration management data, incident and problem management data, stability and resiliency management data, application and infrastructure defect data, risk and security management data, and market data.

6. The method of claim 1 , wherein the proposed change is received in a batch comprising a plurality of proposed changes.

7. The method of claim 1 , wherein the proposed change is incomplete, and further comprising:

requesting additional information necessary to complete the proposed change.

8. The method of claim 1 , further comprising:

monitoring an actual failure and an actual impact of the implemented proposed change; and

scoring the model based on the actual failure and the actual impact.

9. The method of claim 8 , further comprising:

selecting a new model in response to the score being below a predetermined threshold.

10. A system for predictive technology incident reduction, comprising:

a service management system comprising at least one computer processor and executing a change screening module and a change management module;

a learning engine module; and

a change implementation module;

wherein:

the change screening module receives a change record for a proposed change to a computer application or a computer network infrastructure;

the change screening module analyzes the proposed change for an adverse potential cross impact on another computer application or a computer system;

the change screening module predicts a probability of failure and an impact of the adverse potential cross impact on the other computer application or computer system using a model selected by the learning engine module;

in response to a low predicted probability of failure, or a high predicted probability of failure with a low predicted impact:

the change management processor approves the proposed change; and

the change implementation module implements the proposed change; and

in response to a high predicted probability of failure and a high predicted impact, the change management processor rejects the proposed change.

11. The system of claim 10 , wherein the learning engine module selects the model based on a prior implemented change and a failure and an impact of the prior implemented change.

12. The system of claim 10 , wherein the model comprises at least one of a Nave Bayes algorithm, a k-nearest neighbors algorithm, logistic regression, and an ensemble classifier.

13. The system of claim 10 , wherein the learning engine module selects the model by providing test data to a plurality of models, and selects the model based on results of the test data.

14. The system of claim 10 , wherein the predicted probability of failure and the predicted impact of the proposed change are based on at last one of current change data, change history data, configuration management data, incident and problem management data, stability and resiliency management data, application and infrastructure defect data, risk and security management data, and market data.

15. The system of claim 10 , wherein the proposed change is received in a batch comprising a plurality of proposed changes.

16. The system of claim 10 , wherein the change screening module compares at least one of the predicted probability of failure and the predicted impact of the proposed change to a human assessment.

17. The system of claim 10 , wherein the change screening model requests additional information necessary to complete the proposed change.

18. The system of claim 10 , wherein the learning engine module monitors an actual failure and an actual impact of the implemented proposed change and scores the model based on the actual failure and the actual impact.

19. The system of claim 18 , wherein the learning engine module selects a new model in response to the score being below a predetermined threshold.