5. CLAIMS
We claim:
1. A computer implemented system for real-time operational predictive scoring of components and services of an information technology system for forecasting and assessing performance of said components of said information technology system, comprising:
a data pipeline configured to collect, in real-time, a plurality of time series signals comprising multiple metrics corresponding to one or more of health, performance, and functionality of each of said components of said information technology system;
a data store configured to receive a data stream comprising said collected plurality of time series signals, and store said received data stream as individual time series data;
a scoring module, comprising:
an operational predictive score engine configured to calculate an operational predictive score for each of said time series signals of said components of said information technology system corresponding to said individual time series data of each of said plurality of time series signals; and
an operational predictive score roll-up module configured to aggregate said calculated operational predictive score of each of said time series signals of said components of said information technology system into an operational predictive score for each of said components of said information technology system.

2. The system of claim 1, wherein said operational predictive score engine comprises a forecasting engine and a statistical scoring engine.
3. The system of claim 2, wherein said forecasting engine is configured to generate a probabilistic forecast for each of said time series signals of said components of said information technology system, using a plurality of customized machine learning models.
4. The system of claim 3, wherein said statistical scoring engine is configured to calculate said operational predictive score for each of said time series signals of said components of said information technology system, using a statistical model on said probabilistic forecast from said forecasting engine for each of said time series signals of said components of said information technology system.
5. The system of claim 4, further comprising a statistical modeling engine configured to calculate a probability of said plurality of time series signals moving into an erroneous state in near future time, using a statistical model on said probabilistic forecast from said forecasting engine for each of said time series signals of said components of said information technology system, wherein said probability calculated by said statistical modeling engine is used to derive said operational predictive score for each of said time series signals of said components of said information technology system.
6. The system of claim 3, further comprising a training module configured to train said customized machine learning models comprising autoregressive Recurrent Neural Network based forecasting models.
7. The system of claim 6, wherein said training module is further configured to train a common machine learned model for a set of related time series signals.

8. The system of claim 1, wherein said data store is further configured to receive said operational predictive score of each of said time series signals of said components of said information technology system from said operational predictive score engine, and said operational predictive score for each of said components of said information technology system from said operational predictive score roll-up module.
9. The system of claim 1, further comprising an alerting engine configured to monitor a value of said operational predictive score for each of said time series signals and said components of said information technology system, and alert a user when said value falls below a set threshold.
10. The system of claim 1, further comprising a graphical user interface configured to provide a visualization of said operational predictive scores of each of said time series signals and said components to said user, for facilitating one or more of real-time monitoring, historical trend analysis, and system improvement.
11. The system of claim 1, further comprising an operational predictive score quality monitor module configured to monitor a quality and an effectiveness of said operational predictive score of each of said time series signals and said components.
12. The system of claim 6, further comprising automatic triggering of said training module for retraining said customized machines learning models when a quality and an effectiveness of said operational predictive score of any of said time series signals and said components falls below a pre-defined value.

13. The system of claim 1, further comprising calculating an operational predictive score for said services implemented using said components of said information technology system.
14. The system of claim 13, further comprising calculating an operational predictive score for said components in said services implemented using said components of said information technology system.
15. The system of claim 14, further comprising a root cause analysis module configured to decompose said operational predictive score for any of said components of said information technology system into a responsibility matrix when said operational predictive score for any of said components of said information technology system deteriorates during a user transaction journey within said services implemented using said components of said information technology system, to locate one or more of said components that are contributing to said deterioration.
16. A computer implemented method for real-time operational predictive scoring of components and services of an information technology system for forecasting and assessing performance of said components of said information technology system, comprising:
providing a computer implemented system comprising a data pipeline, a data store, a scoring module comprising an operational predictive score engine and an operational predictive score roll-up module, a statistical modeling engine, a training module, an alerting engine, a graphical user interface, an operational predictive score quality monitor module, and a root cause analysis module;
collecting, in real-time, by said data pipeline, a plurality of time series signals comprising multiple metrics corresponding to one or more of

health, performance, and functionality of each of said components of said information technology system;
receiving, by said data store, a data stream comprising said collected plurality of time series signals, and storing said received data stream as individual time series data;
calculating, by said operational predictive score engine of said scoring module, an operational predictive score for each of said time series signals of said components of said information technology system; and
aggregating, by said operational predictive score roll-up module of said scoring module, said calculated operational predictive score of each of said time series signals of said components of said information technology system into an operational predictive score for each of said components of said information technology system.
17. The method of claim 16, wherein said operational predictive score engine comprises a forecasting engine and a statistical scoring engine.
18. The method of claim 17, further comprising generating by said forecasting engine a probabilistic forecast for each of said time series signals of said components of said information technology system, using a plurality of customized machine learning models.
19. The method of claim 18, further comprising calculating by said statistical scoring engine said operational predictive score for each of said time series signals of said components of said information technology system, using a statistical model on said probabilistic forecast from said forecasting engine for each of said time series signals of said components of said information technology system.

20. The method of claim 19, further comprising calculating, by said statistical modeling engine, a probability of said plurality of time series signals moving into an erroneous state in near future time, using a statistical model on said probabilistic forecast from said forecasting engine for each of said time series signals of said components of said information technology system, wherein said probability calculated by said statistical modeling engine is used to derive said operational predictive score for each of said time series signals of said components of said information technology system.
21. The method of claim 18, further comprising training, by said training module, said customized machine learning models comprising autoregressive Recurrent Neural Network based forecasting models.
22. The method of claim 21, further comprising training, by said training module, a common machine learned model for a set of related times series signals.
23. The method of claim 16, further comprising receiving, by said data store, said operational predictive score of each of said time series signals of said components of said information technology system from said operational predictive score engine, and said operational predictive score for each of said components of said information technology system from said operational predictive score roll-up module.
24. The method of claim 16, further comprising monitoring, by said alerting engine, a value of said operational predictive score of each of said time series signals and said components, and alert a user when said value falls below a set threshold.

25. The method of claim 16, further comprising providing, on said graphical user interface, a visualization of said operational predictive score of each of said time series signals and said components to said user, for facilitating one or more of real-time monitoring, historical trend analysis, and system improvement.
26. The method of claim 16, further comprising monitoring, by said operational predictive score quality monitor module, a quality and an effectiveness of said operational predictive score of each of said time series signals and said components.
27. The method of claim 22, further comprising automatic triggering of said training module for retraining of said customized machine learning models when a quality and an effectiveness of said operational predictive score of any of said time series signals and said components falls below a pre-defined value.
28. The method of claim 16, further comprising calculating an operational predictive score for said services implemented using said components of said information technology system.
29. The method of claim 28, further comprising calculating said operational predictive score for said components in said services implemented using said components of said information technology system.
30. The method of claim 29, further comprising decomposing, by said root cause analysis module, said operational predictive score for each of said components of said information technology system into a responsibility matrix when said operational predictive score for any of said components of said information technology system deteriorates during a user transaction journey within said services implemented using said components of said information technology

system, to locate one or more of said components that are contributing to said deterioration.