AGENTIC AI-POWERED AUTONOMOUS SOFTWARE ENGINEERING FRAMEWORK FOR AUTOMATED CODE GENERATION AND DEBUGGING.
This paper discusses the disruptive nature of agentic AI-based autonomous systems in software engineering, specifically the automated code generation and debugging. The main aim is to analyze the role of agentic AI-based systems, who combine autonomy, reasoning as well as adaptive learning in improv...
| Publicado en: | Scientific Culture Vol. 12; no. 1, Part 1; pp. 2816 - 2823 |
|---|---|
| Autores principales: | , , , , , |
| Formato: | Literature Review |
| Publicado: |
University of the Aegean
2026
|
| Materias: | |
| Acceso en línea: | Ver este registro en EBSCOhost |
| fields | @attributes: recordID: 1 pdfLink: plink: https://search.ebscohost.com/login.aspx?direct=true&db=hlh&AN=192213736&site=ehost-live header: @attributes: shortDbName: hlh uiTerm: 192213736 longDbName: Humanities International Complete uiTag: AN controlInfo: bkinfo: jinfo: jid: 24080071 I6HU jtl: Scientific Culture issn: 24080071 maglogo: N pubinfo: dt: 2026 vid: 12 iid: 1, Part 1 pid: 47715 pub: University of the Aegean artinfo: ui: 192213736 10.5281/zenodo.121126204 ppf: 2816 ppct: 7 formats: tig: atl: AGENTIC AI-POWERED AUTONOMOUS SOFTWARE ENGINEERING FRAMEWORK FOR AUTOMATED CODE GENERATION AND DEBUGGING. aug: au: Maheshkar, Jaykumar Ambadas Vankayala, Himaja Jakkula, Vamshi Krishna Raj, Leslie Daniel Khedekar, Pratik Laheri, Rohit affil: Group Application Manager-Sr, USBancorp, USA. Sr. Application Developer, Baltimore City Office of Information & Technology (BCIT), Maryland, USA. Senior Software Engineer, Independent Researcher, IEEE member, USA. Senior Cloud Application Architect, Amazon Web Services, Inc., Florida, United States. Data Scientist, Independent Reseacher, USA. Software Engineer, Tech Mahindra, Texas, USA. su: Multiagent systems Code generators Software frameworks Automatic control systems Software engineering Intelligent agents Machine learning sug: subj: Multiagent systems Code generators Software frameworks Automatic control systems Software engineering Intelligent agents Machine learning keyword: Agentic AI AI Autonomous Code Debugging Governance Multi-Agent Resilience Scalability Systems ab: This paper discusses the disruptive nature of agentic AI-based autonomous systems in software engineering, specifically the automated code generation and debugging. The main aim is to analyze the role of agentic AI-based systems, who combine autonomy, reasoning as well as adaptive learning in improving efficiency, accuracy, resilience, governance, and scalability within the contemporary development setting. In the study, the secondary research method is used, as the data synthesizes available academic evidence in the form of journals, surveys, and systematic reviews. Secondary data collection included thematic identification, screening and extraction of suitable literature and thematic analysis was used to classify results into coherence themes that included efficiency gains, debugging accuracy, adaptive resilience, governance mechanisms, and multi-agent scalability. Findings show that language models based on transformers and multi-agent systems enhance the generation of code and decrease the technical debt and increase productivity. Debugging frameworks based on machine learning improve precision by identifying some of the latent patterns of error and reducing the time spent in resolution. Adaptive learning strategies enhance resilience by using fault-tolerant designs and self-repairing algorithms, and the governance systems maintain accountability, transparency, and moral control. The collaboration between agents also facilitates scalability through distributed orchestration and cooperative solutions. All these results make agentic AI a software engineering paradigm shift that integrates both automation and governance-conscientious autonomy to provide robust, explainable, and future-proof solutions. The paper ends by stating that technical resilience, ethical protection, and adaptive resilience should be combined together to achieve the maximum potential of autonomous software engineering. pubtype: Academic Journal doctype: Literature Review src: R language: English refInfo: copyright: @attributes: flag: Y dt: @attributes: year: 2026 holdings: @attributes: islocal: N |
|---|