Chaitanya Rahalkar
Software Security Engineer, Block
As a security professional at Block Inc focused on building scalable cloud-native data pipelines and detection systems, I bring valuable perspective on innovative security solutions that truly work at scale.
My background including security engineering at Praetorian, Meta, and now Block Inc gives me broad exposure to diverse security challenges and solutions across different industries and scales. Having worked on projects from fuzzing platforms to payment systems security, I can evaluate submissions from multiple technical angles while understanding their practical business impact.
I'm passionate about fostering innovation in cybersecurity and believe my experience building scalable security systems combined with my academic research background would help identify truly groundbreaking contributions to the field. The opportunity to recognize excellence while connecting with other security leaders aligns with my commitment to advancing cybersecurity practices across the industry.
All Sessions by Chaitanya Rahalkar
Track A - The Linguistics of Large Language Models: What Your AI's Mistakes Reveal
Front ClassroomWhen GPT-4 writes "I'll send you the attachment later" (without any ability to send attachments) or ChatGPT claims it can "see" an image that isn't there, what's really happening? This talk dives into the fascinating patterns of AI hallucinations, exploring how linguistic analysis of AI errors provides unique insights into how these models actually work. Through live examples, we'll examine common patterns of LLM mistakes and what they reveal about the underlying architecture and limitations of current AI systems. Key Points: 1. Common patterns in AI hallucinations and their linguistic roots 2. The disconnect between capability claims and actual abilities 3. How context windows influence AI behavior 4. Understanding prompt injection through linguistic analysis 5. Real-world examples of AI linguistic patterns 6. What these patterns tell us about future AI development
Track B - Navigating AI Security: Protecting Your Organization in the Era of Generative AI
Track BDesigned for security professionals and technology leaders, this session requires no deep technical knowledge but assumes familiarity with basic security concepts and enterprise AI use cases. We'll use recent incidents as case studies to demonstrate how to develop security policies that balance innovation with risk management.