Stop AI From Going Rogue - Build Secure RAG Pipelines
JANUARY 23, 2025 | 9AM PT | 12PM ET
Prevent OWASP Top 10 Threats Like Prompt Injection
RAG (Retrieval Augmented Generation) pipelines enable your LLMs to access real-time and proprietary data for more accurate responses. In this workshop, we'll build a RAG chat app that uses sensitive documents in Google Drive, and top it off with identity + access control and learn how to defend against prompt injection, data leaks, and malicious actors.
***Note: If you'd like to follow along with the hands-on demos, please join the workshop from your laptops.
Agenda:
- Build a secure RAG chat app that uses real-time, proprietary data safely
- Defend AI apps against prompt injections and other OWASP top 10 LLM threats
- Integrate identity and access control to protect sensitive data in RAG pipelines and AI agents.
.png?width=100&height=100&name=Untitled%20design%20(30).png "Untitled design (30)")
Pranav Shikarpur is a Developer Advocate at Pangea, focused on helping developers build secure by design apps. He is also one of the early contributors to learnprompting.org - the largest open-source guide on prompt engineering. Previously, he honed his expertise in data security as a developer advocate at Thales, educating developers on the importance of data encryption and key management. Having been the founder of a seed-funded ed-tech startup in the past, he has a wealth of experience in app development in both startup and enterprise environments.
