Context
Modern AI coding tools allow developers and agents to generate hundreds of lines of code in seconds. However, our standard processes for managing software—like Git feature branches and pull requests—were created over two decades ago when writing code by hand was a slow process. Because code generation is now nearly instant, the main bottleneck in software engineering has shifted from writing code to reviewing, testing, and integrating it.
Motivation
While recent studies have examined code quality and static analysis findings in individual pull requests, the broader impact on repository dynamics remains underexplored. Key factors—such as how AI agents affect branch lifespans, reviewer effort, merge friction, and long-term issue traceability—lack empirical answers. Software engineering teams need rigorous, data-driven evidence rather than marketing claims to design modern governance models for AI-native workflows.
Goal
The primary goal of this thesis is to empirically analyze how AI-assisted development alters branching dynamics, review friction, and post-merge maintenance compared to human-authored baselines. Using datasets of agentic pull requests (e.g., the AIDev dataset), the student will analyze branch- and PR-level metrics across open-source repositories.
The outcomes of this thesis include:
- An empirical benchmark quantifying the impact of AI adoption on branch lifespans, review latency, merge friction, and downstream defects.
- A traceability assessment operationalizing and evaluating issue–branch–PR semantic linkage in AI-assisted workflows.
- A final report providing evidence-based recommendations and process model guidelines for AI-native branch governance.
- A contribution toward a potential scientific publication.
Requirements
- Programming skills in Python (including data analysis tools such as Pandas)
- Strong familiarity with Git, GitHub APIs, and software development workflows
- Interest in empirical software engineering, repository mining, and software quality
Pointers
- AIDev Dataset: Dataset of agent-attributed pull requests and repository histories. ()
- SAIL Research: AI Teammates in Software Engineering. ()
- Recent Literature: Studies analyzing AI coding agent adoption, development velocity, static-analysis warnings, and code maintainability (e.g., ACM Digital Library studies on AIDev).
Contact
Main Contact: Dr. Nitish Patkar