nao Labs

nao Labs is a France-based open-source software company behind nao, an analytics agent builder designed around the emerging discipline of context engineering. The product lets data teams construct an agent's context like a file system: databases, documentation, queries, repositories, semantic layers, and configuration files are assembled into a structured context that an AI agent can use to answer analytics questions over a company's data. The company positions nao as the analytics agent built for context engineering, with the core message that agent reliability depends on context, and that context should be engineered deliberately rather than left to chance. nao is distributed as 100% open source and is operated through a command-line workflow. Commands such as `nao init`, `nao sync`, and `nao test` let users create a file-system context for their agent, pull context automatically from existing sources such as databases, dbt repositories, Looker, and Notion, and measure context reliability through unit tests of questions to SQL. The context can include warehouse schemas, metrics and models, business definitions, example queries, rules, docs, tools, and MCPs, with no stated limit on what can be added. A chat UI can then be deployed so that anyone in an organization can run analytics on the data. Beyond the open-source product, nao Labs offers services to help companies implement analytics agents. These include full AI data agent setup, context engineering advisory, context engineering training, and custom personalization of AI data agents. The company describes working with clients to design their context layer, wire their stack, and launch AI data agents that business users can actually use, targeting reliability levels of 80% or higher while optimizing cost and performance. Deployment can be in the customer's own infrastructure, including a nao UI and Slack or Teams bot. The company is led by co-founders Claire, who serves as CEO, and Christophe, who works in engineering, alongside team members such as Mateo and Stan in engineering. nao Labs is presented as an open-source analytics software company. Its users include data leaders at organizations such as LiveKit, Everyday, Finfrog, and Qare, who report using nao to absorb ad hoc product questions, support department heads and executives in analytics, and deliver fast answers during high-stakes meetings. The company maintains a users wall of fame to showcase these stories and invites teams to share how they use nao.

nao Labs emerged as an open-source analytics software company building nao, an analytics agent builder centered on context engineering. The company's public materials describe a product philosophy in which agent reliability depends on context, and where context should be engineered as deliberately as software. Rather than treating an AI analytics agent as a black box, nao exposes the context layer as a file system that teams can create, synchronize, and test. The product workflow is organized around three core commands. `nao init` creates the file-system context for an agent, allowing users to add data, metadata, rules, docs, tools, and MCPs without a stated limit. `nao sync` pulls context automatically from existing sources such as databases, dbt repositories, Looker, and Notion. `nao test` measures and monitors context performance through unit tests of questions to SQL, so teams can improve reliability over time. A chat UI can be deployed so that anyone can run analytics on the data, and the agent can also be surfaced through Slack or Teams bots. nao Labs complements its open-source product with professional services. According to its services page, the company helps clients implement nao end to end, including building a context layer from warehouse, repository, and internal business tool context, setting up agent evaluation with unit tests, and improving agents to reach 80%+ reliability with optimal cost and performance. It also offers context engineering advisory, training sessions on core context engineering concepts, and custom personalization covering security enforcement, UI customization, and context source customization. An intensive one-day hackathon format is available for a specific scope, producing a context layer, agent evaluation with unit tests, and a data agent ready to deploy. The company is led by co-founders Claire (CEO) and Christophe (engineering), with Mateo and Stan also in engineering roles. Its users include data leaders at LiveKit, Everyday, Finfrog, and Qare. Testimonials describe nao absorbing ad hoc product questions, enabling department heads and CEOs to rely on it for analytics, and delivering a 'wow moment' when a CTO answered a question from a Medical Director during a steering committee using a few prompts in nao. nao Labs maintains a users wall of fame to feature these stories and invites teams to submit their own.

Open-source analytics agent builder focused on context engineering for reliable AI data agents.

• nao analytics agent builder

  Open-source platform for building analytics agents with engineered context.

• nao CLI

  Command-line workflow with nao init, nao sync, and nao test for creating, synchronizing, and testing agent context.

• nao chat UI

  Deployable chat interface for anyone to run analytics on data.

• Slack and Teams bot

  Deploy the nao agent in Slack or Teams within the customer's own infrastructure.

1. 1nao | Open Source Analytics Agent Buildergetnao.io
2. 2Services | naogetnao.io
3. 3Teamgetnao.io
4. 4Users wall of fame — data teams using nao | naogetnao.io
5. 5nao Labswikidata.org