nao

The nao product provides an open-source analytics agent designed for data teams to interact with data using natural language and to build context-driven analytics across chat interfaces and live dashboards. It enables self-hosting, integrates with common data tools, and emphasizes a context layer that governs agent reliability. The tool supports collaboration across an organization by letting multiple users query and explore data through an agent-driven interface.

The nao platform operates as a framework to build and deploy analytics agents. It uses a file-system like context for the agent, supports data stacks, and offers a UI for chat interactions to run analytics on data. It also provides a testable, self-hosted environment where teams can manage context and agent behavior.

Users interact with nao via chat to ask questions, see visualizations, and explore data in dashboards and stories. The product highlights natural language queries, native data visualizations, and sharing of insights through stories and live dashboards. It emphasizes transparency in the agent’s reasoning and sources, with feedback workflows to improve results.

nao emphasizes open-source foundations, with self-hosting options and control over data, security, and access. It provides enterprise capabilities like SSO, row-level security, and governance features in addition to its self-hosted/open-source core. The project aims to enable data teams to act as context engineers, building reliable, auditable analytics agents.