Databricks

Prototype AI agents using AI Playground for MCP testing.

Securely access data with managed MCP servers for AI.

AI-powered transaction categorization for scalable, accurate data processing.

Uses reinforcement learning for improved SQL translation accuracy.

Curated marketplace for certified data products by domain.

AI-powered visual pipeline builder for data engineering tasks.

Build AI agents and applications using Mosaic AI platform.

Provides detailed tracing for GenAI applications in production.

New experience for business users to leverage AI and data.

Serve machine learning features and models with Lakebase.

Fully-managed Postgres database for AI and data applications.

Improves agent quality using natural language feedback guidance.

Supports over 250,000 QPS for enterprise AI applications.

Managed, auto-scaling GPUs for AI workloads in beta.

Scalable vector search with lower cost for RAG applications.

Faster, multi-modal AI functions for SQL data transformations.

Redesigned for GenAI with agent observability and prompt versioning.

Build high-quality agents optimized for data and cost.

Ensures validated applications meet quality requirements before production.

Manages prompt versions with Git-style tracking and testing.

Captures application snapshots for tracking changes and improvements.

Web interface for collecting structured expert feedback on AI outputs.

Research-backed evaluation for measuring GenAI quality systematically.

Unified platform for AI experimentation and observability.

Evaluates AI agents' performance on Databricks.