Nao is an AI powered data IDE for analysts, engineers, and scientists to write SQL, Python, or dbt workflows, preview changes, catch issues early, and deploy confidently. It connects directly to your warehouse and understands your schema so you can build faster, fix fewer bugs, and maintain trust in your data. Build data pipelines, launch quality checks, run analytics, and collaborate across teams without context switching. Think of it as your AI teammate built specifically for modern data work.
I respect Nao's effort to automate data tasks. The agentic coding area is established, yet agentic analytics and data science development are lacking. I am happy to see a company advancing this area.
What needs improvement
In addition to writing DBT models, I'd like the editor to have an intuitive way to plan and create analysis-related tables, data models, and visualizations. This will speed up and simplify the analysis works significantly.
Hi Product Hunt, I’m Claire, co-founder of nao Labs! 👋🏻
nao is the AI data IDE I wish I had when I was working in data.
2 years ago I was head of data at sunday. I was trying to keep up with the speed of the business, helping my team avoid breaking changes in prod, while trying to save them time for the interesting analytics stuff.
But it was tough. Our tools just seemed so unadapted to our work. We kept switching from the IDE to the warehouse console and manually applying data model changes in the BI tool. I kept reconnecting extensions until I just gave up.
Then AI came!
But once again, AI coding tools were thought for developers, not for data people. They handle code context very well, but not the variety of context you need for data work: data schema, documentation, data stack, business definitions.
So one year ago, when Christophe and I started thinking about how to make data tooling more efficient for data people, it became clear: we wanted to create the best place to work on data with AI.
nao is the AI data IDE, designed specifically for data workflows. It’s a fork of VSCode, directly connected to your warehouse. And the AI agent has all the context of your code + metadata + data stack and business context - all in a secure local setup.
We released our first version 6 months ago, so I can tell you a bit about what our users love:
It’s all in one. Anyone in the data team - technical or not - can easily connect their data and stack integration in one click.
One prompt from data schema to a full data pipeline with data quality checks.
One prompt for deep-dive analysis. nao plans analytics to run, runs all checks, and provide a full analytics summary.
nao is still in beta and there’s much more coming. Our goal is to be fully integrated across your entire data stack, end-to-end — from data engineering to data science to analytics. We believe that for AI to be adopted in analytics, it must first be adopted and trusted by the data team itself.
We have a free trial and free version - give it a try and let us know what you think!
Happy data vibing,
Report
@claire_gouze Congrats on the launch. I like how clean everything looks... the demo video, landing page, gallery images. :)
@claire_gouze This is one of the best PH launch comments I've read in a while. No buzzwords - just a real story about a real problem. The journey from "I gave up on extensions" to building the solution yourself is an inspiring one. Congratulations on the launch!
@nuseir_yassin1 With nao as it's a local IDE supporting git workflows you can handle versioning acrross your pipelines with git (or whatever other scm you want to use)!
hey Product Hunt, Christophe here, co-founder of nao Labs!
I've been working in the data industry for more than 10 years and I think that data people finally deserve a tool that is made for them. AI is completely reshaping how software, data and analytics is made. On the software side we've witnessed major breakthrough, but on the data side we are still working with clunky UIs, juggling between tools and changing code without even knowing what are the impact of the changes.
With nao we want to change this, we believe data people can have better tools in which they can build, fix and explore data while using AI to do their daily job faster than before and focus on high value-added tasks.
You can try nao for free! It's the best way to work on data using AI.
@mateolbs@claire_gouze Hey Claire and Mateo, congrats on the launch of nao. If you're ever in need of a marketing analyst to help in user adoption of this great product, I'll love to discuss on how my experience can contribute to your success.
Report
Congrats on the launch @claire_gouze and @bleff ! @nao hits a real gap in data tooling. I’ve spent years architecting, building and operating pipelines across analytics and finance, and the friction is always the same: context and code lives in too many places and tools (warehouse, BI, docs, code), so changes are slow and risky.
An IDE that’s natively "data-aware" (schema, lineage, business definitions) and connects directly to the warehouse is the right abstraction.
My experience so far has been fantastic:
Connected to my warehouse and generated a dbt model + tests from tables in minutes; lineage preview caught a join issue early.
Ran an ad‑hoc analysis via prompt; the agent planned the checks and produced a solid summary with links back to sources.
Environment switching worked cleanly for dev → staging; diffs were readable and prevented drift.
Curious about two things:
Do you plan to continue enhancing the SQL query experience? I'm still switching between nao and DataGrip for some tasks.
Do you plan to offer BYOK in the PRO plan? I would love to use our own Bedrock keys for the Claude models.
@benjamin_sicard thanks for the detailed feedback, great to see you could do so many things with it.
Yes of course SQL experience is one of our key obsessions -> anything on Datagrip you're missing on nao?
For now, BYOK is for enterprise plan, let's get in touch to see what could be fit for you!
Report
Hello @claire_gouze@bleff - Any plan to squeeze the whole set of .naorules into a single folder? Having multiple naorules files on the root directory of a repo with hundreds of dbt models and yml files looks suboptimal to us
heyyy @_aneema , this is a great question! At the moment we have a single .naorules in the root folder, but as rules are growing in size we are considering moving everything inside a single folder with easy to use selectors.
Just to be sure would you like to have a single folder at the top with selectors or would you prefer being able to add .naorules in subfolders of your dbt project?
@_aneema for now you can use one single .naorules so that it's just one file at the root of your folder. But we have on our roadmap to have the agent understand multiple files of .md in your repo!
Report
@claire_gouze@bleff Thanks for getting back to me. I think Nao rules should live in /.naorules (place in the root directory) with extension specific sub-rules. For example, /.naorules/sql-rules.txt stores formatting rules for .sql files, /.naorules/python-rules.txt for Python files, etc. (.txt is a placeholder, not the actual extension I would assign).
I also think this could save a nontrivial amount of tokens in API calls: when processing only .sql files, Nao passes only sql-rules.txt to the LLM instead of the entire generic .naorules file containing rules for all extensions. More manageable for data teams: analytics engineers maintain rules for their files, data engineers maintain rules for theirs, data scientists their, etc Hope it makes sense, happy to help if you have any further questions :)
nao
Thank you for your review Chau, that's very great to read! We're working on a way to make nao even better for exploration and analytics. Stay tuned!