The page is a product landing page for Bonnard, a software tool for agent-native analytics and data visualization. It contains no editorial discussion of human rights. Structural signals show mild positive alignment with privacy (default tracking denial), economic accessibility (free tier), and support for cultural/scientific creation through data analysis tools. The content is commercially focused with no explicit human rights framing.
My brother Alex and I have been building in the data space for over a year. We started with AI-powered report creation, triggered analysis, deep-dive automation and some other stuff for data teams, finance, and ops. Custom products we'd build for each client. We had some traction, but the AI could never deliver accurately on the last 10%. Insane amounts of context and prompt engineering went into it, but clients didn't want it unless they had full trust in the numbers. Which is, frankly, impossible.
We figured out a few things.
1. If someone else builds your analysis, you're going to redo it anyway. That's just how people work. If I handed you a presentation someone else made, you'd probably rebuild it from scratch. And with numbers, it's even worse because you're going to pick them apart regardless.
2. Unless you're an analyst iterating on SQL, there is no point putting code in front of a business user. So they can learn and debug SQL? lol.
3. Self-serve failed because it was the UI that was self-serve and not the actual data. Business users do not want to learn another UI low-code tool. The world is moving in the opposite direction.
Business users should just use the tools they already know: ChatGPT/Codex, Claude Desktop, whatever. Those tools are already insanely powerful at delivering cheap UI. All you need to do is make the data self-serve, fast and reliable for agents. It feels pretty obvious that every business with data to operationalise should let people do whatever the hell they need to do with it to get the job done faster, right? It's not the data team's problem anymore cause the UI is cheap. The only thing that the data team needs to do with it is keep enriching and maintaining the self-serve data layer. They can then actually get time to do the impactful stuff.
So we built Bonnard. Data teams (Analytics Engineer, Data engineers or whoever has been handed the AI delivery project) go from data model to self-serve in about 5 minutes, all through the CLI. Zero UI on the consumer side unless they build it themselves.
Would love to get your feedback! Will hang in the comments if you have any questions.
I've spent so much of my career as a data gopher, so I really appreciate this. It always works much, much better when people have direct access to data... but success always relied on getting business users to be a bit more technical.
I built a version of this in my org as a slack bot. It has access to BigQuery and I worked with it to document and define the key tables, so when a user asks a question it has a "library" to figure out how to answer it. Then it will respond directly in Slack (plus respond to f/up questions in the thread) and if it's a bigger/more detailed question it can create the report in Metabase and send a link.
One thing I don't have is the governance aspect, so I may give this a spin to see if this is a good middle layer.
Looking forward to getting your feedback if you get a chance to play around! It takes less than a few minutes to get set up and starting experimenting with it. Deploying coded dashboards is super good fun as well - deploying correct data in seconds is pretty mindblowing how difficult it was to get things up and running in the past
The page includes Google Tag Manager script with default consent set to 'denied' for ad_storage, ad_user_data, ad_personalization, and analytics_storage.
Inferences
The default denial of tracking suggests the site considers user privacy in its implementation.