How we built a city-scale event aggregation platform using LLMs for every variable, judgment-intensive task, onboarding, parsing, classification, and enrichment and the architectural pattern that
Explore how we keep AI and machine learning systems secure and respectful of user privacy. As models become more powerful, protecting sensitive data and building trustworthy systems is more important than ever. An AI feature is also a new way for data to leave your company: prompts get logged, embeddings sit in a vector database nobody owns, and a helpful agent reads a document it should never have been given. In this category, you'll find posts that cover challenges, strategies, and tools for safeguarding AI workflows.
