AI-Driven Code Review: Balancing Automation with Developer Autonomy

Authors: Chandra Prakash Singh

DOI: https://doi.org/10.5281/zenodo.14615572

Short DOI: https://doi.org/g8x3p9

Country: USA

Full-text Research PDF File:   View   |   Download

Abstract: The use of AI-driven tools in software development has revolutionized tasks such as code generation, debugging, and performance optimization. Large language models (LLMs) like ChatGPT, GitHub Copilot, and Codeium provide developers with robust assistance, streamlining the coding process. However, understanding their strengths and limitations remains a critical area of study. This paper evaluates these tools' performance on LeetCode challenges across varying difficulty levels, analyzing success rates, runtime efficiency, memory usage, and error-handling capabilities. GitHub Copilot excelled in generating accurate solutions for simpler tasks, while ChatGPT demonstrated superior debugging and memory efficiency. Codeium, while promising, struggled with complex problems. This study offers actionable insights for developers and researchers aiming to optimize the integration of AI tools into their workflows.

Keywords: ChatGPT, GitHub Copilot, Codeium, LeetCode, AI in Software Development, Code Generation, Debugging, Error Handling, Efficiency Analysis

Paper Id: 232006

Published On: 2025-01-06

Published In: Volume 13, Issue 1, January-February 2025

Cite This: AI-Driven Code Review: Balancing Automation with Developer Autonomy - Chandra Prakash Singh - IJIRMPS Volume 13, Issue 1, January-February 2025. DOI 10.5281/zenodo.14615572

Share this