System and a method for optimizing sprint-based tasks in agile methodology

Patent metadata

Field	Value
Publication number	US-12505408-B2
Application number	US-202318212385-A
Country	US
Kind code	B2
Filing date	Jun 21, 2023
Priority date	Apr 11, 2023
Publication date	Dec 23, 2025
Grant date	Dec 23, 2025

Abstract

A system for optimizing sprint-based tasks implemented in an agile methodology, the system including a memory and a processor configured to execute a task optimization engine to receive input data including story point data associated with historical user stories captured during previous sprints, where the input data is analyzed to determine a first feature dataset, where a timeseries dataset is determined for forecasting unplanned task for upcoming sprint based on analysis of the input data associated with unplanned task of the previous sprints and a dataset associated with attributes is determined based on the input data associated with the previous sprints for story point data of an upcoming sprint, and where the datasets are combined to generate a persistent identifier for sprint capacity buffer data values to optimize sprint-based tasks in agile methodology.

First claim

We claim:

1 . A system for optimizing sprint-based tasks implemented in an agile methodology, the system comprising:
a memory storing program instructions; a processor executing program instructions stored in the memory and configured to execute a task optimization engine configured to: receive, via one or more Application Programming Interfaces (APIs), input data from one or more Application Lifecycle Management (ALM) tools, the input data comprising story point data associated with historical user stories captured during previous sprints, wherein the input data is analyzed to determine a first feature dataset; determine a reference threshold value associated with the story point data such that the first feature dataset corresponding to the story point data is in a pre-defined range; determine a second feature dataset from the input data, the second feature dataset comprises data values obtained based on the previous sprints; determine a third feature dataset by analyzing the second feature dataset to determine correlated axis values associated with the data values of the previous sprints to determine defect counts associated with the previous sprints; determine a timeseries dataset for forecasting unplanned task for upcoming sprint based on analysis of the input data associated with unplanned task of the previous sprints; determine a dataset associated with attributes based on the input data corresponding to the previous sprints for story point data of an upcoming sprint, wherein one of the attributes includes determining repeatable tasks within repeatable stories by employing a nested algorithm based-model, and wherein a similarity threshold parameter is employed such that clusters of user stories are formed and tasks associated with the user stories that are determined to be above the similarity threshold are marked employing clustering techniques; and combine the datasets to generate a persistent identifier for sprint capacity buffer data values to optimize sprint-based tasks in agile methodology, wherein the system is configured for optimizing sprint based tasks in agile methodology by reducing uncertainties and enhancing productivity.

2 . The system as claimed in claim 1 , wherein the task optimization engine that receives the input data from an input data unit, the input data unit is in communication with the ALM tools via the APIs.

3 . The system as claimed in claim 1 , wherein the task optimization engine comprises an estimation effort analysis unit configured to analyse the input data to determine the first feature dataset that represents efforts involved in executing story points employing Inter Quartile Range (IQR) value, lower or upper boundaries and outliers relating to tasks associated with the previous sprints, wherein any overlap in the IQR value of effort across story points indicates poor story point estimation that is visualised using box plots via a User Interface (UI).

4 . The system as claimed in claim 3 , wherein the estimation effort analysis unit determines the reference threshold value employing a statistical model, wherein the reference threshold value relates to completed user stories with story points such that an effort estimation is in the predetermined range of a standard deviation.

5 . The system as claimed in claim 3 , wherein the estimation effort analysis unit is configured to identify a story point drift towards a higher story point value or lower story point value, where the story point drifts towards the lower story point value indicates higher productivity and the story point drift towards higher story point indicates less productivity.

6 . The system as claimed in claim 3 , wherein the estimation effort analysis unit is configured to determine an overshoot value, wherein user story with the effort value that lies outside single standard deviation is marked as the overshoot value.

7 . The system as claimed in claim 1 , wherein the task optimization engine comprises a pre-processing unit configured to determine the second feature dataset, the second feature dataset comprises:
first data values associated with new terms identified in user stories of an upcoming sprint that are not present in the previous sprint; second data values associated with number of user stories; third data values associated with number of developers; fourth data values associated with number of Quality Assurance (QA) personnel; fifth data values associated with complexity indicated by distribution of user stories across story points; sixth data values associated with module wise count of user stories; seventh data values associated with change in team members captured in terms of incremental change compared to the previous sprints, wherein the second feature data is configurable depending on data availability.

8 . The system as claimed in claim 1 , wherein the task optimization engine comprises a defect prediction unit configured to employ Principal Component Analysis (PCA) on the second feature dataset to identify the correlated axis values.

9 . The system as claimed in claim 8 , wherein the defect prediction unit is configured to execute a regression model employing a XGBoost regressor based technique to train the regression model based on the second feature dataset computed from the previous sprints along with the defect counts.

10 . The system as claimed in claim 8 , wherein the defect prediction unit is configured to compare user stories of an upcoming sprint with defect counts using cosine similarity and store comparison data as a fourth feature dataset.

11 . The system as claimed in claim 8 , wherein the defect prediction unit is configured to identify one or more attributes that contribute to defects, wherein the one or more attributes are identified via a correlation operation.

12 . The system as claimed in claim 8 , wherein the defect prediction unit is configured to compress the second feature dataset into a correlated feature dataset that is achieved via PCA analysis which enables deriving a new set of feature values that represent values from underlying features through use of eigen values and eigne vectors.

13 . The system as claimed in claim 1 , wherein the task optimization engine comprises a forecasting unit configured to determine the timeseries dataset for forecasting unplanned task for upcoming sprint and store as a fifth feature dataset.

14 . The system as claimed in claim 13 , wherein the forecasting unit is configured to execute a Generative Adaptive Model (GAM) on the input data for forecasting the unplanned task.

15 . The system as claimed in claim 1 , wherein the task optimization engine comprises a risk estimation unit configured to determine the attribute including velocity data that represents over allocation or under allocation at a team member level, wherein the velocity data is compared with story point data of the upcoming sprint and the comparison data is stored as a sixth feature dataset.

16 . The system as claimed in claim 15 , wherein the risk estimation unit is configured to determine the attribute including new user stories and store it as a seventh feature dataset, and wherein the new stories are identified by using One-Class SVM model, the model is trained on the user stories from the historical input data and anomalies found in the user stories in the upcoming sprint.

17 . The system as claimed in claim 15 , wherein the risk estimation unit is configured to execute named entity recognition on the input data to identify anomalies in user stories in the upcoming sprint using Natural Language Proce

Assignees

• Cognizant Tech Solutions India Pvt Ltd

Inventors

• Kulkarni Dattaprasad
• Patil Hemant
• Kumaran Rajeswari
• Grover Shikha
• Singh Lucky

Classifications

• G06F11/302

  where the computing system component is a software system · CPC title

• G06Q10/06313

  Resource planning in a project environment · CPC title

• G06F11/3688

  for test execution, e.g. scheduling of test suites · CPC title

• G06F8/71

  Version control (security arrangements therefor G06F21/57); Configuration management · CPC title

• G06Q10/06311 Primary

  Scheduling, planning or task assignment for a person or group · CPC title