"noobs united team on IBM Bob 2.0 hackathon Hackathon"
Team Idea
Debugging failed tests wastes significant developer time — manually tracing failures to root causes, reproducing issues, and verifying fixes one by one. Our project uses IBM Bob 2.0's multi-agent capabilities to automate this workflow. A triage agent runs the test suite and clusters related failures; a root-cause agent diagnoses why each cluster fails; a repair agent proposes fixes; a verification agent reruns tests to confirm fixes work without regressions. Independent failure clusters are processed in parallel using Bob's subagents. We demonstrate this on a sample codebase with seeded bugs, measuring before/after time-to-resolution, tests fixed per session, and regression rate — showing how multi-agent orchestration reduces debugging effort and turnaround time.
Team Members
.png&w=256&q=75)
Mohammed Ayman Rayhan
Ayman_Rayhan
.png&w=640&q=75)