part of Mosaic AI Mosaic AI works for all types of AI — classical ML, deep learning, and generative AI. MLOps + LLMOps Implement CI/CD processes and automation across all of your AI workloads Prepare Data & Vectors Prepare data & features with native tools Serve Applications Serve models into real-time apps, with monitoring and alerts Build & Evaluate Models Train or fine-tune custom models, and prompt engineer pre-trained models Serve Data & Vectors Serve structured data and vector databases as real-time APIs Governance Govern data and AI models together under Unity Catalog Data Platform Build on top of your open Lakehouse, powered by Delta Lake
Build & Evaluate Models Prepare Data & Vectors MLOps + LLMOps Serve Data & Vectors Governance Data Platform RAG is part of Mosaic AI Mosaic AI capability (AI) Lakehouse capability (Data + AI) Mosaic AI works for all types of AI — classical ML, deep learning, and generative AI. MLflow Asset Bundles (DABs) CI/CD support SQL Workflows Delta Live Tables Notebooks Model Serving AI Functions Models from SQL Lakehouse Apps Lakehouse Monitoring Notebooks w/ ML Runtime AutoML Fine Tuning Models in Marketplace MPT LLaMA2 AI Playground MLflow Track & Evaluate Feature Serving Vector Search Function Serving Model Registry in Unity Catalog Unity Catalog Feature Store in Unity Catalog Delta Tables Structured Data Files (Volumes) Unstructured Data RAG-related update in Dec. 2023