NOVA: A New Open-Source Framework for Orchestrating AI Agents
NOVA is an upcoming open-source system designed to help users build and manage teams of multiple AI agents. It uses local orchestration to enhance data privacy and lower the costs of using premium AI models.
NOVA (Networked Orchestration of Virtual Agents) is a framework that connects specialized AI agents using the Model Context Protocol (MCP). By allowing users to run the orchestration layer on their own hardware through self-hosting, it provides a cost-effective alternative to expensive all-in-one AI platforms. The system is built to handle complex workflows by assigning specific tasks to the most appropriate agents. This modular design enables builders to use cheaper models for simple tasks and reserve expensive models only for critical steps, significantly reducing overall token usage and operating expenses.
Key points
- Connects multiple specialized agents to work together as a unified team.
- Uses the Model Context Protocol (MCP) to link agents with local tools and databases.
- Supports self-hosting to keep data secure and eliminate high cloud subscription fees.
- Provides fine-grained control over model selection to optimize for both performance and cost.
Quick term guide
- open-source
- Software whose code is shared publicly so others can inspect, use, or change it.
- system design
- The plan for how different parts of a website or app work together to handle users and data.
- orchestration
- Coordinating multiple AI agents or steps to run in a specific order or in parallel to complete a task
- Model Context Protocol (MCP)
- A standard way to connect AI models to outside data sources and tools.
- Model Context Protocol
- A shared standard that defines how AI assistants connect to and use outside tools and services
- self-hosting
- Running the software on your own server instead of relying fully on an outside service.
- self-host
- To run a website, app, or service on your own server instead of using a hosted provider.
- subscription
- A pricing model where you pay a fixed amount of money every month for access.