OnDemand AI Agents is a cloud platform for building and running agent-based AI systems using a RAG approach. It’s designed for teams that want to roll out intelligent automation quickly without managing complex infrastructure.
Agent platform and centralized control
At its core is a decentralized operating system for agents with a single control center to configure agents, manage permissions, and connect them to internal systems.
- Centralized management for agents and access rights
- BYOM support (bring your own model)
- Connect multiple LLMs through a unified integration layer
Visual flow builder and reusable agents
OnDemand includes a visual Agent Flow Builder for assembling automation workflows from blocks rather than full custom development. You can start from preconfigured agents or build your own by combining tools, APIs, and data sources.
- Drag-and-drop workflow building
- Library of ready-to-use agents
- Custom agents built from tools, APIs, and data
Data, documents, and testing before production
The platform focuses on secure enterprise data integration, including dynamic APIs, internal/external storage connections, and file processing for RAG queries. A built-in Playground helps test models and agent configurations before deployment.
- Secure data and document retrieval for RAG
- LLM Playground for evaluation and iteration
- Scales from pilots to distributed agent networks
