May 2, 2026
Notebook? More like note-brawl
Show HN: Mljar Studio – local AI data analyst that saves analysis as notebooks
Your AI data helper stays on your laptop, but commenters are already fighting over the receipts
TLDR: MLJAR Studio wants to be a private AI helper for data work that runs fully on your own laptop and saves everything as notebooks. Commenters liked the human-review safety angle but immediately argued over whether notebooks are trustworthy and whether the product is truly different from existing tools.
A new project on Hacker News is pitching a very 2026 fantasy: an AI data analyst that lives entirely on your own computer. No sending company secrets to the cloud, no mystery black box, and every step saved as a notebook you can review later. On paper, that sounds like catnip for anyone who wants help exploring spreadsheets, building prediction tools, and turning their findings into shareable apps without handing over sensitive data.
But the comment section? Instant drama. The loudest eyebrow-raise came from people side-eyeing the product’s big promise of reproducibility. One commenter basically said: wait, you’re fixing unreproducible AI chat by saving it in notebooks... the same notebooks people famously mess up with hidden changes and out-of-order clicks. Ouch. Another crowd went straight for the comparison game, asking why not just use Deepnote or a Jupyter MCP Server setup with Claude, where the assistant can already write and run notebooks. Translation: is this a real new product, or just a slicker wrapper around stuff people already have?
The most serious takes were less snarky and more nervous: if software starts making data-based recommendations, somebody still needs to sanity-check it before it burns money. Zillow’s very public modeling disaster got name-dropped like a cautionary ghost story. So the vibe is clear: people love the privacy angle, but they’re demanding proof this isn’t just "chatbot with a lab coat".
Key Points
- •MLJAR Studio is described as a fully local AI data analysis and machine learning tool that runs on a user’s computer.
- •The product allows users to ask questions in natural language and receive locally executed Python code and results.
- •The article says its AI agent can iteratively improve notebooks and run machine learning experiments to search for better models.
- •Users can get notebook assistance for code writing, transformations, and visualizations while retaining control over execution.
- •The article says notebooks can be converted into web applications using the open-source Mercury framework and hosted on a team’s own infrastructure.