On Tuesday, Monte Carlo introduced Agent Observability, a new feature that integrates data and AI observability, enhancing the development of reliable AI applications. Previously, the platform focused on data management processes but lacked insights for AI initiatives. This unification allows customers to monitor data integrity, retrieval from knowledge sources, and AI output quality—critical for effective AI performance. With over 80% of AI projects failing to reach production due to poor data quality and trust issues, Agent Observability aims to bridge this gap. It utilizes LLM-as-judge monitors to evaluate AI outputs, mapping them back to source data for easy problem identification. Monte Carlo CEO Barr Moses emphasized that customer feedback guided this development, addressing challenges in trusted data for AI. Looking ahead, Monte Carlo plans to enhance its observability capabilities and strengthen partnerships with AI governance platforms to ensure comprehensive data integrity and compliance, positioning itself as a leader in the data observability market.
