As Large Language Models (LLMs) transition from conversational tools to autonomous AI agents capable of high-stakes, multi-step actions, the reliance on generalized, default safety alignment from model providers has become insufficient and often hazardous. This session addresses the critical need for application-specific, custom guardrails necessitated by persistent jailbreak vulnerabilities, the high cost of achieving effective fine-tuning alignment, and the challenge of enforcing proprietary organizational compliance (e.g., contextual PII protection or specific regulatory mandates unique...
Aditya Gautam
Machine Learning Tech Lead
Meta
Aditya Gautam is a Machine Learning Tech Lead at Meta, where he spearheads the development of generative AI architectures, multimodal agents, and next-generation recommendation systems. His research covers topics including real-world multimodal understanding, multi-agent systems, and LLM evaluation, and he has published in top-tier conferences such as ICWSM and WSDM. His leadership in the field has been featured in high-profile interviews with Nasdaq and major technical publications like Marktechpost and Uniquely Wired, where he has shared defining insights on the future of autonomous agents, AI agent applications, and high-ROI multi-agent systems. A sought-after speaker at premier global stages—including Silicon Slopes Summit, WSDM, and the Databricks Data + AI Summi, Aditya’s recent work explores the scaling of AI agent architectures, system efficiency, and the broader landscape of enterprise adoption for multimodal systems. Committed to advancing the scientific community, he serves as a peer reviewer for venues like NeurIPS, ICML, AAAI, and ACM, and has served on the program committee for ECML-PKDD.