Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{201717,
        author = {Karan S Khedkar and AnilKumar J Kadam},
        title = {A MULTI-AGENT AUTONOMOUS SOFTWARE ENGINEERING FRAMEWORK USING LLM-BASED AGENTS},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {12},
        pages = {5284-5292},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=201717},
        abstract = {As software applications continue to grow in complexity, developers often spend significant time on planning, structuring, and writing repetitive code components. Existing AI coding assistants mainly provide autocomplete suggestions or generate small code snippets, but they usually lack the ability to handle complete software development workflows in an organized and autonomous manner. To overcome these limitations, this work presents a Multi-Agent Autonomous Software Engineering Framework that converts user requirements written in natural language into structured software projects. The framework uses multiple AI agents, where each agent is assigned a specific responsibility such as project planning, architecture generation, and source-code development. A workflow-based orchestration approach is used to coordinate communication between agents and manage the execution sequence. The system also supports structured task generation, recursive execution flow, and tool-based file operations for creating and managing project files automatically. The proposed framework is implemented using LangGraph, ReAct-based agents, and large language models to support intelligent reasoning during software generation. Initial testing performed on different software generation tasks such as CRUD applications, API-based systems, and utility projects shows that the framework can generate organized project structures and complete implementation steps with consistent performance. The study demonstrates how multi-agent AI systems can support the development of autonomous and scalable software engineering solutions beyond traditional code suggestion tools.},
        keywords = {Autonomous software engineering, multi-agent systems, code generation, large language models, LangGraph, ReAct agents, AI-based coding systems, workflow orchestration, intelligent software development, autonomous coding framework, software automation, agentic AI systems},
        month = {May},
        }