Detecting and Mapping Code Defects Introduced by Generative AI Using CodeLogic & Neo4j
summary
CodeLogic maps code as a Neo4j graph and uses CypherQL plus combined artifact and runtime scanning to flag breaking changes that generative AI edits can introduce before merge, not after. Live demo shows GitHub Copilot refactor breaking a remote REST endpoint, change requests reveal cross‑app impact, and the release assistant simulates library upgrades so maintainers can plan. PR annotations push Slack alerts to reviewers. Internal use prevents regressions and enabled a Spring 6 migration by one engineer, proving the approach works at scale.
AI-written summary · about
chapters (13)
- 0:00Introduction and company overview
- 1:12The butterfly effect in code
- 3:41Why relational databases failed
- 5:48Software intelligence platform explained
- 7:34AI challenges and generative AI impact
- 9:21Demo: Refactoring with GitHub Copilot
- 12:12Impact analysis in CodeLogic
- 15:24Release Assistant for library updates
- 17:10Real-world use case with Spring 6 upgrade
- 18:54Testing implications and benefits
- 20:22Data collection methods and agents
- 22:08Future AI integrations and predictions
- 24:53Code fearlessly and Q&A
related talks
24mGitHub Copilot Agents + Sonar | The Blueprint for Agentic Development
41mCode Security Reinvented Navigating the era of AI
58mSecuring Code in the Age of AI
46mBreaking AI: Live coding and hacking applications with Generative AI
29mOne-Click Code Fix: Securing Code Using AI
23mAI-Generated Code: Who Is Responsible for the Bugs?
47mExploring a complex codebase with AI
30mWhat Are the Security Risks Associated with AI-Generated Code? 🤔