Description:FIELD OF THE INVENTION
The present invention relates to the field of task and project management systems. The present invention specifically relates to a system and method for identifying, analyzing, and merging duplicate tasks across multiple projects and teams. The system leverages state-of-the-art tuned Large Language Models (LLMs) in combination with deterministic systems, such as context graphs, to automate the identification, relationship analysis, and merging of duplicate tasks. By integrating with source systems like Git, JIRA, and other workplace tools, this invention aims to improve productivity and consistency across teams.
BACKGROUND OF THE INVENTION
With the increasing complexity of managing tasks across multiple projects and teams, the identification and resolution of duplicate tasks have become critical challenges in modern project management. Duplicate tasks often arise due to overlapping responsibilities, lack of centralized task tracking, and disparate project management systems. Addressing these duplicates manually requires significant effort in identifying, analyzing, and resolving redundancies, often leading to inefficiencies, inconsistencies, and missed deadlines.
Existing approaches to managing duplicate tasks are fragmented and insufficient, relying heavily on manual processes or basic matching methods that lack the sophistication to analyze task intent, dependencies, and priorities comprehensively. These methods often fail to provide accurate identification of duplicates, do not account for complex relationships between tasks, and overlook the need for synchronized updates across related tasks. Consequently, organizations face increased operational overhead, reduced productivity, and diminished team collaboration.
There is a growing need for a robust and intelligent system that can seamlessly integrate with existing project management tools to identify, analyze, and merge duplicate tasks with precision. Such a system must address critical aspects like intent matching, dependency analysis, and automated updates, providing a unified and scalable solution for task management across diverse workflows.
US7483841B1 describes a system for processing data using machine learning models to improve task handling within a computational environment. While it offers advancements in task execution, it primarily focuses on pre-defined workflows and static training, lacking a mechanism for real-time model updates or integration with evolving data sources. This invention does not address continuous dynamic task allocation or real-time fine-tuning of models in response to changing tasks, making it less suitable for environments where ongoing adaptation to new data and requirements is crucial.
US10095999B2 presents a method for evaluating tasks and allocating resources using machine learning models. While it provides a structure for task management and resource distribution, it does not incorporate continuous updates or real-time feedback loops to enhance model performance or task handling. The system also lacks a mechanism for ensuring that tasks are assigned to the most suitable models based on context and task-specific requirements, limiting its applicability for dynamic environments that require continuous fine-tuning.
OBJECTS OF THE INVENTION
The primary objective of the invention is to provide a system and method for identifying, analyzing, and merging duplicate tasks across multiple projects and teams.
Another objective of the invention is to provide a system and method to automate the identification, relationship analysis, and merging of duplicate tasks by leveraging state-of-the-art tuned Large Language Models (LLMs) in combination with deterministic systems, such as context graphs.
A further objective of the invention is to provide a system and method that extracts tasks from various project management and version control systems, analyzes the tasks using both LLMs and deterministic models, and generates unique and duplicate task lists.
Yet another objective of the invention is to provide a system and method that propagates the changes made to the base task to other related tasks, ensuring uniformity and reducing manual effort.
An additional objective of the invention is to provide a system and method that leverages advanced LLMs and deterministic model to reduce redundancy, ensures consistency, and enhances productivity.
SUMMARY OF THE INVENTION
Before the present invention is described, it is to be understood that the present invention is not limited to specific methodologies and materials described, as these may vary as per the person skilled in the art. It is also to be understood that the terminology used in the description is for the purpose of describing the particular embodiments only and is not intended to limit the scope of the present invention.
The present invention discloses a system and method for identifying and merging duplicate tasks across multiple projects, addressing challenges related to task redundancy, dependency tracking, and cross-project consistency. Central to the system is a connect to source systems module, which retrieves tasks and associated data from workplace tools like JIRA, Git, and similar platforms, ensuring a standardized dataset for comprehensive analysis.
The system includes an analyze tasks with LLMs module that utilizes fine-tuned language models to interpret task descriptions, extract actionable insights such as intent, approach, and dependencies, and generate structured summaries. A context graph construction module maps task relationships and dependencies using advanced graph-based techniques, offering a holistic view of task connections and overlaps.
A key feature of the system is the duplicate task identification module, which employs a hybrid methodology combining LLMs and deterministic scoring to identify duplicate tasks based on semantic similarity, shared dependencies, and team roles. The merge tasks module resolves these duplicates by selecting a base task using a priority-based framework that considers urgency, deadlines, and workload distribution, ensuring consistency and efficiency in task management.
The system also includes a propagate changes module to automate the application of updates from the base task to all related tasks, maintaining uniformity and reducing manual intervention. Furthermore, a dashboard module provides an intuitive interface for reviewing and approving duplicate identification and merging decisions, enhancing user control and transparency.
This invention improves task management by reducing redundancy, ensuring consistency, and fostering collaboration across teams. The system's scalable and adaptable architecture supports a variety of project management tools and scenarios, offering a comprehensive solution for streamlining workflows and enhancing productivity in modern workplace environments.
DETAILED DESCRIPTION OF THE INVENTION
Before the present invention is described, it is to be understood that this invention is not limited to methodologies described, as these may vary as per the person skilled in the art. It is also to be understood that the terminology used in the description is for the purpose of describing the particular embodiments only and is not intended to limit the scope of the present invention. Throughout this specification, the word “comprise”, or variations such as “comprises” or “comprising”, will be understood to imply the inclusion of a stated element, integer or step, or group of elements, integers or steps, but not the exclusion of any other element, integer or step, or group of elements, integers or steps. The use of the expression “at least” or “at least one” suggests the use of one or more elements or ingredients or quantities, as the use may be in the embodiment of the invention to achieve one or more of the desired objects or results. Various embodiments of the present invention are described below. It is, however, noted that the present invention is not limited to these embodiments, but rather the intention is that modifications that are apparent are also included.
The present invention outlines an advanced system designed to identify and consolidate duplicate tasks across various projects by integrating fine-tuned language models (LLMs) with deterministic systems. The primary objective of this system is to analyze tasks from different platforms, detect duplicates, and merge them while retaining their respective priorities and dependencies. By utilizing a modular design, the system ensures seamless integration with a variety of workplace tools, enhancing task management efficiency. The invention is a multi-step system that extracts tasks from various project management and version control systems, analyzes the tasks using both LLMs and deterministic models, and generates unique and duplicate task lists. Duplicate tasks are merged based on priority and dependencies. Changes made to the base task are propagated to other related tasks, ensuring uniformity and reducing manual effort. The system also integrates visual relationship graphs for better dependency tracking.
The system consists of the following key modules: connection to source systems module, data extraction module, task analysis module, context graph construction module, duplicate identification module, priority based merging module, generation task list module, propagation of changes module. In the present invention, each module plays an integral role in achieving the system’s goals, as outlined below:
Connection to Source Systems module:
The connection to source systems module connects to various source systems, such as Git, JIRA, and others, via APIs to retrieve all tasks and issues. The connection layer is designed to support a wide range of workplace tools using a modular plugin architecture. It also retrieves essential task details such as title, description, priority, and status, facilitating smooth integration with diverse project management environments by using API.
Data Extraction module:
The Data Extraction module processes the retrieved task data and stores it in a standardized structured database. This module normalizes task information, such as task ID, title, description, priority, and other relevant attributes, ensuring that tasks from different platforms are stored in a consistent format. The normalization process guarantees uniformity across the dataset, allowing subsequent analysis and task management to be performed without inconsistencies.
Task analysis module using tuned LLMs:
Task analysis module uses a tuned large language model LLM to analyze each task, leveraging fine-tuned models trained on a proprietary dataset consisting of diverse task descriptions, project management logs, and annotated dependencies. This ensures that the LLM is optimized for extracting actionable insights specific to workplace environments. The LLM generates a structured summary for each task, including: intent of the task, details of the approach and change summary.
Context Graph Construction module:
A context graph is built to map relationships and dependencies between tasks using a hybrid algorithm that combines heuristics and graph theory principles. The system employs specific techniques like cosine similarity for task description comparison and temporal mapping to establish dependencies based on task creation timelines. This approach ensures a robust and efficient graph structure tailored for workplace task management. Nodes in the graph represent tasks, and edges represent dependencies, such as prerequisites or shared objectives.
Duplicate task Identification module:
This module traverses the context graph and uses a hybrid approach combining the LLM and a deterministic scoring system to identify duplicate or similar tasks based on similar intents scored using semantic similarity metrics, shared dependencies determined via context graph analysis and overlapping team members analyzed through role and responsibility alignment.
Priority -based merging module:
This module checks the priority of duplicate tasks and selects the highest-priority task as the base using a proprietary evaluation framework. This framework considers not only the assigned priority levels but also contextual factors such as deadline urgency, resource availability, and team workload distribution to ensure optimal task selection. Changes from this base task are propagated to all related tasks through a robust propagation mechanism. This mechanism employs a dependency-aware update algorithm, ensuring that all related tasks receive consistent updates while avoiding conflicts. The system validates changes against the context graph to maintain task dependencies and uses version control techniques to resolve discrepancies in real-time, ensuring seamless integration across the task network.
Generation of Unique and Duplicate Task Lists module:
The system generates two lists: one with unique tasks and another with duplicate tasks. These lists are presented to the user via a dashboard for review and approval. For example, the dashboard highlights duplicate tasks in red, allowing the user to inspect and confirm merging decisions.
Propagation of changes module:
Once a duplicate task is merged, the system ensures that changes made to the base task are automatically propagated to all related tasks. For example: If the “Optimize API structure” task requires updating a code module, the same update is applied to all related tasks.

Examples:
Use Case 1: Cross-Team Collaboration
Two teams working on separate projects independently create tasks to “Optimize API endpoints.” The system identifies these as duplicates and merges them into a single high-priority task.
Use Case 2: Dependency Resolution
A task to “Update front-end forms” depends on a “Database schema update” task. The context graph highlights this dependency, ensuring changes are synchronized.

This invention provides a comprehensive solution for managing duplicate tasks across projects and teams. By leveraging advanced LLMs and deterministic models, the system reduces redundancy, ensures consistency, and enhances productivity.

While considerable emphasis has been placed herein on the specific elements of the preferred embodiment, it will be appreciated that many alterations can be made and that many modifications can be made in preferred embodiment without departing from the principles of the invention. These and other changes in the preferred embodiments of the invention will be apparent to those skilled in the art from the disclosure herein, whereby it is to be distinctly understood that the foregoing descriptive matter is to be interpreted merely as illustrative of the invention and not as a limitation.
, Claims:We claim,
1. A system and method for identifying and merging duplicate tasks across multiple projects
characterized in that
the system comprises of a connection to Source Systems module that is configured to integrate with project management and version control tools, to retrieve tasks and issues; a data extraction module that normalizes and structures the retrieved task data for further processing; a task analysis module that utilizes tuned large language models to analyze task content and generates structured summaries including task intent, approach, and dependencies; a context graph construction module that creates a visual representation of task relationships and dependencies using graph theory principles and temporal mapping; a duplicate identification module that uses a hybrid LLM and deterministic scoring system to identify duplicate tasks which are based on task similarity, dependencies, and team members; a priority-based merging module that merges duplicate tasks by selecting a base task based on priority, dependency analysis, and additional contextual factors; a task list generation module that generates and displays unique and duplicate task lists to the user via a dashboard for review and approval; and a change propagation module that ensures updates to the base task are automatically propagated to all related tasks across the system;
the method for identifying and merging duplicate tasks, comprising the steps of:
a. retrieving and normalizing task data from source systems;
b. analyzing tasks using tuned LLMs to extract task intent, approach, and dependencies;
c. constructing a context graph to map relationships and dependencies;
d. identifying duplicate tasks based on similarity, dependencies, and team member overlap;
e. merging duplicate tasks based on priority;
f. propagating changes to related tasks;
g. generating unique and duplicate task lists for review.
2. The system and method as claimed in claim 1, wherein the source system connection module integrates with workplace tools through a modular plugin architecture.
3. The system and method as claimed in claim 1,wherein the task analysis module uses fine-tuned LLMs which are trained on a proprietary dataset to analyze task descriptions and generate actionable insights.
4. The system and method as claimed in claim 1, wherein the context graph construction module uses techniques such as cosine similarity and temporal mapping to establish task dependencies.
5. The system and method as claimed in claim 1, wherein the duplicate identification module flags the duplicate tasks based on task similarity, shared dependencies, and overlapping team members.
6. The system and method as claimed in claim 1, wherein the priority-based merging module selects the base task considering factors such as task priority, deadline urgency, and team workload.
7. The system and method of claim 1, wherein the change propagation module uses a dependency-aware update mechanism to ensure that changes to the base task are reflected in all related tasks.
8. The system and method of claim 1, the step of analyzing tasks using tuned LLMs includes generating summaries that capture the intent, approach, and dependencies of each task.
9. The system and method of claim 1, wherein the step of identifying and merging duplicate tasks uses a combination of semantic analysis, dependency analysis, and priority evaluation to optimize task merging.