Research Scientist · CSIC · Madrid
Biomedical Engineer and Master in Robotics with a passion for AI and LLMs.
Explore my universe
Star chart
Every star is something I built, published, learned or won. Hover to read, click to travel.
Latest transmissions
Recent work
Tangent
Semantic peer-matching network for early-career researchers.
POIROT
Open-source protocol for failure detection and attribution in LLM multi-agent systems.
POIROT: Interrogating Agents for Failure Detection in Multi-Agent Systems
Orchestrating Large Language Models into Multi-Agent Systems (LLM-MAS) has unlocked remarkable reasoning capabilities, yet emergent failures and hallucinations that resist characterisation block their deployment in safety-critical domains — a gap made legally untenable by emerging AI regulation. Existing evaluation paradigms share a common flaw: centralised judgment creates single points of failure and demands domain-specific expertise. Here we present POIROT, a protocol that repurposes a system's own agents as its diagnostic layer, leveraging the epistemic diversity already present in the architecture. Across evaluated settings, POIROT outperforms single-LLM evaluator baselines, with gains that scale with problem complexity (OR = 1.60, p = 0.008), agent count, and fault dimensionality, persisting under compound fault conditions. These results demonstrate that safety oversight need not be externalised: the agents executing a role carry sufficient collective intelligence to audit it. We release POIROT as an open-source library alongside BLAME, a benchmark for fault attribution in safety-critical multi-agent systems.