AI-Driven Code Generation:Challenges and Opportunities

Thesis title: AI-Driven Code Generation:Challenges and Opportunities
Author: Öztorun, Mustafa
Thesis type: Diploma thesis
Supervisor: Černý, Jan
Opponents: Potančok, Martin
Thesis language: English
Abstract:
The recent emergence of large language models (LLMs), followed by AI-assisted coding assistants, revolutionized the traditional software development industry by introducing AI-assisted coding tools like GitHub Copilot, Cursor, Windsurf, and Replit. These tools enhance productivity, reduce repetitive tasks by assisting developers, and support developing and maintaining comprehensive applications. Despite their growing popularity and adoption, their effectiveness, limitations, weaknesses, and real-world impact remain underexplored, particularly in complex development scenarios. This research investigates the capabilities, challenges, and opportunities of AI Coding Assistants throughout the development life cycle by developing a comprehensive web application with the aid of the said tools as a case study from the perspective of a solo developer working without the support of a dedicated development team. The study evaluates key dimensions such as AI assistant task coverage capabilities, prompt intent understanding and robustness, debugging, testing capabilities, and documentation. A comparative analysis of tools like Replit, Cursor, Windsurf, and GitHub Copilot highlights their strengths and weaknesses. By demonstrating the practical use cases of AI Coding Assistants, this study provides insights into their potential to enhance software development and support solo developers while emphasizing the need for careful supervision.
Keywords: AI-assisted coding; GitHub Copilot; Cursor; Code Generation; Large Language Models (LLMs); Replit; Productivity; Windsurf; Artificial Intelligence (AI); Generative Artificial Intelligence; Integrated Development Environment (IDE)
Thesis title: AI-Driven Code Generation:Challenges and Opportunities
Author: Öztorun, Mustafa
Thesis type: Diplomová práce
Supervisor: Černý, Jan
Opponents: Potančok, Martin
Thesis language: English
Abstract:
The recent emergence of large language models (LLMs), followed by AI-assisted coding assistants, revolutionized the traditional software development industry by introducing AI-assisted coding tools like GitHub Copilot, Cursor, Windsurf, and Replit. These tools enhance productivity, reduce repetitive tasks by assisting developers, and support developing and maintaining comprehensive applications. Despite their growing popularity and adoption, their effectiveness, limitations, weaknesses, and real-world impact remain underexplored, particularly in complex development scenarios. This research investigates the capabilities, challenges, and opportunities of AI Coding Assistants throughout the development life cycle by developing a comprehensive web application with the aid of the said tools as a case study from the perspective of a solo developer working without the support of a dedicated development team. The study evaluates key dimensions such as AI assistant task coverage capabilities, prompt intent understanding and robustness, debugging, testing capabilities, and documentation. A comparative analysis of tools like Replit, Cursor, Windsurf, and GitHub Copilot highlights their strengths and weaknesses. By demonstrating the practical use cases of AI Coding Assistants, this study provides insights into their potential to enhance software development and support solo developers while emphasizing the need for careful supervision.
Keywords: AI-assisted coding; Productivity; Artificial Intelligence (AI); Generative Artificial Intelligence; Large Language Models (LLMs); Windsurf ; GitHub Copilot; Cursor; Replit; Code Generation; Integrated Development Environment (IDE)

Information about study

Study programme: Information Systems Management/Data and Business
Type of study programme: Magisterský studijní program
Assigned degree: Ing.
Institutions assigning academic degree: Vysoká škola ekonomická v Praze
Faculty: Faculty of Informatics and Statistics
Department: Department of Information Technologies

Information on submission and defense

Date of assignment: 11. 10. 2024
Date of submission: 24. 6. 2026
Date of defense: 20. 8. 2026
Identifier in the InSIS system: https://insis.vse.cz/zp/89906/podrobnosti

Files for download

    Last update: