JunctionX Korea 2026
ARGUS
Monitoring an AI:GO agent squad during benchmark problem solving.
Overall Winner
1st Place, Lablup & FuriosaAI joint track
JunctionX Korea 2026
August 2026
Overview
Our team, Demo Day Care, built an AI:GO agent squad for benchmark problem solving and ARGUS, a dashboard for monitoring its execution.
My work
- Model-role experiments. Contributed to experiments with model roles in the agent squad.
- Validation tests. Contributed tests for configuration normalization, candidate artifacts, submission, and runtime validation.