Session Information
Keeping Dependencies Up to Date with AI: A Practical Journey to Better Code Quality and Faster Development in Vue Projects
Updating dependencies may seem like a minor task, but it's often nerve-wracking. With the rapid evolution of AI, however, that upgrade experience is starting to change in very real ways.
In this session, we'll explore how we used AI tools like Cursor and Claude Code to drive the upgrade of a real Vue codebase in a production service. Key highlights include:
* Semi-automating the entire process from impact analysis to code transformation and test generation
* Integrating AI into the QA process to cross-check specifications and test cases in real time
* Visualizing project buffers and constraint slack to anticipate hidden risks
* Measuring tangible improvements across multiple metrics—build times, bundle size, and cyclomatic complexity
* Creating a feedback loop by logging AI suggestions for ongoing review and retraining
We'll share what got easier, what didn't go as planned, and how it *felt* to make this shift—giving you a candid look at the real-world impact of bringing AI into the upgrade process.
sayn0
Keeping Dependencies Up to Date with AI: A Practical Journey to Better Code Quality and Faster Development in Vue Projects
Updating dependencies may seem like a minor task, but it's often nerve-wracking. With the rapid evolution of AI, however, that upgrade experience is starting to change in very real ways.
In this session, we'll explore how we used AI tools like Cursor and Claude Code to drive the upgrade of a real Vue codebase in a production service. Key highlights include:
* Semi-automating the entire process from impact analysis to code transformation and test generation
* Integrating AI into the QA process to cross-check specifications and test cases in real time
* Visualizing project buffers and constraint slack to anticipate hidden risks
* Measuring tangible improvements across multiple metrics—build times, bundle size, and cyclomatic complexity
* Creating a feedback loop by logging AI suggestions for ongoing review and retraining
We'll share what got easier, what didn't go as planned, and how it *felt* to make this shift—giving you a candid look at the real-world impact of bringing AI into the upgrade process.