nao

nao is the Analytics Agent built for context engineering. Build and control your agent's context, deploy to business users

nao lets data teams build analytics agents whose context they control, then hand those agents to business users through a chat interface. The core idea is context engineering: you create the agent's context as a file system, with folders for databases, docs, queries, repos and semantics, plus a config file and a rules file. Anything can go in, including data, metadata, business definitions, example queries, dbt repositories, Notion pages, tools and MCPs, and a single init command scaffolds the structure.

Once the context is in place, nao deploys a chat UI so anyone in the company can ask questions about the data, and the site describes an evaluate step for keeping agents reliable. It works with your existing data stack, with a BigQuery warehouse shown in the sample configuration, and the whole project is 100 percent open source with code on GitHub. Coding and development teams get docs, services and user stories to learn from.

The software is open source; a separate pricing page exists, but the plans are not detailed in the excerpt. Run the init command and start with a single database to see how the agent answers.

Category: Coding & Development. Pricing: free. Visit website

Alternatives to nao in Coding & Development

• LaunchRepo — Directory-launch boilerplate your own AI coding agent runs for you.
• SubmitMyStartup — Find where to submit your startup: MCP server + planner
• iCoderz Solution — On-Demand Apps & Software Solutions
• ERIUS PHONE — Lets AI agents test real Android apps on a phone via API
• CPFHub.io — Brazilian CPF lookup API with an MCP server for AI agents
• HProxy MCP — Free proxy list, proxy checker and IP lookup for AI agents