tools
DataRobot Agentic AI
Enterprise platform for building, deploying, monitoring, and governing AI agents across cloud, on-premises, hybrid, and air-gapped environments.
DataRobot Agentic AI helps development and operations teams create agent workflows, connect models, tools, data sources, vector stores, and AI components, then evaluate and deploy them.
It includes code-first development, templates, CLI workflows, monitoring, tracing, guardrails, compliance testing, approval workflows, and infrastructure options including serverless, cloud, on-premises, and hybrid deployments. Agentic AI capabilities are premium; pricing requires contacting DataRobot.
Features
- Build agent workflows with templates for CrewAI, LangChain, LlamaIndex, and other frameworks
- Connect LLMs, AI tools, datasets, vector stores, embeddings, and re-rankers
- Evaluate and compare workflows with playgrounds, metrics, and LLM-as-a-judge testing
- Deploy agents through DataRobot's CLI, serverless infrastructure, or supported GPU environments
- Monitor agents with tracing, lineage, custom metrics, service health, and drift detection
- Apply guardrails, moderation, approval workflows, and automated compliance reports
- Integrate external tools and data sources through open architecture, connectors, APIs, and MCP
- Run deployments in cloud, on-premises, hybrid, private-cloud, sovereign, and air-gapped environments
Use cases
- Build production agents for enterprise workflows
- Analyze data across databases, tables, and warehouses
- Create retrieval pipelines for text, images, and PDF documents
- Generate forecasts, summaries, emails, and customer-facing content
- Route requests across LLMs and deploy AI components on NVIDIA infrastructure
- Test agent behavior against prompt injection, scope escalation, and multi-turn attacks