May 2
Loading video player...
Chaitanya Gunupudi, senior cloud platform engineer and AI research practitioner at the University of Maryland, will talk about Building Secure DevSecOps Architectures for AI and LLM Workloads. AI models and LLM-powered features are shipping faster than ever, but most teams are still bolting security on at the end of the pipeline. This session walks through how to design a practical DevSecOps architecture specifically for AI workloads: from secure data pipelines and model training to hardened inference endpoints and continuous monitoring in production. We will cover how to integrate security checks into CI/CD and MLOps, protect against common AI threats (data leakage, model poisoning, and prompt abuse), and build an audit-ready workflow that satisfies both engineering and compliance needs. Attendees will leave with a concrete reference architecture, example controls, and a checklist they can adapt for their own stack whether they’re deploying fine‑tuned models, using managed LLM APIs, or running in a hybrid cloud environment " https://gdg.community.dev/events/details/google-gdg-hudson-presents-hvtech-virtual-building-secure-devsecops-architectures-for-ai-and-llm-workloads/