Evaluation of Agentic System // Aditya Gautam // Agent Hour

Aditya is a seasoned AI expert and thought leader at the nexus of AI Integrity, recommendation systems, and LLM-powered agents, focusing on building trustworthy, efficient AI at scale. As a Machine Learning Technical Lead for Integrity at Meta, he architects large-scale AI systems to improve ranking algorithms, combat misinformation, and improve user engagement. He previously served as a founding engineer for a Computer Vision startup within Google’s prestigious Area 120 incubator.

Aditya is quite active in Generative AI community with him being a sought after speaker, panelist, and interviewee, frequently sharing novel insights on agentic system evaluation, LLM cost optimization on industry podcasts and at premier summits like the Databricks Data + AI Summit 2025, Marktechpost, AI agent conference, Analytics Vidhya, MLops Community and other. His expertise, particularly on Generative AI and agent misinformation, has been featured in major media articles, including the Daily Herald and Marktechpost. His recent research presented at ICWSM 2025 offers a blueprint for a multi-agent system for the misinformation lifecycle. Dedicated to maintaining high standards, he serves as an Ethics Reviewer for NeurIPS 2025 and reviewer several papers for top-tier conferences like ICML, AAAI, ACM among others.