One-on-One with Swayanshu Shanti Pragnya
summary
Ten years of AI research and a current data lead role drive work on security, data governance and interpretability while collaborating with University of Colorado Denver and Stanford. Uses Hugging Face models to build agents but focuses on what training data gets fed and on creating secure model architectures. Runs a comparative neuroscience study measuring neural responses with and without AI assistance and promotes harness engineering—tools to amplify human judgment rather than replace it—while prototyping defensive algorithms.
AI-written summary · about
related talks
30mIntroduction to AI Safety. Let's be ready before it's too late!
26mSafeguarding AI: How to Manage AI and its Potential Risks
41m#define AI Engineer
06mTransitioning AI from the lab to real-life applications
33mEngineering AI Applications: A Conversation with AI Leaders
46mCan you trust your AI?
48mMastering the AI Shift: Adaptive Strategies for Software Evolution
36mModus OperandAI: Practical Security for Artificial Intelligence