Saloni Garg builds and secures machine learning systems at Adobe. Instead of just deploying LLM apps, she focuses on breaking them—testing RAG pipelines and agents for vulnerabilities like prompt injection, data leakage, and tool abuse.
She is currently writing a book on Retrieval-Augmented Generation for De Gruyter, exploring how to build and maintain these architectures safely. Alongside her production work, she holds an AI patent and has published multiple papers with IEEE.