Let Hex's agent abstract away the most tedious parts of data work, so you focus on the new, the novel, and the gnarly. Hex brings the best of Claude Sonnet 4 directly to our powerful notebooks where thousands of data teams are already working.
Hey Product Hunt! I’m Olivia, a product manager at Hex. Today, I’m excited to introduce the Notebook Agent.
For the last few months, data people have been watching as software engineers have been getting first-class agentic tools to do their work, stuck copying and pasting trying to keep up. We built the Notebook Agent to change that.
Now you can get the best of agentic capabilities in Hex. Just ask a question, get to a first draft in minutes, and iterate on any complex analysis with full control over the agent’s edits.
Why Hex? Because the agent lives right where you work — connected to your projects, files, and warehouse tables. It has the context and tools to understand your analysis and can generate SQL, Python, native visualizations, and Markdown summaries.
We’re already seeing folks use it to get started faster with ad hoc exploration and supercharge deep-dive analysis. If you try it out, let us know what you think!
Love how this agent lives right in my notebook—no more context switching just to draft some SQL. Can't wait to see how this changes my exploration workflow. Congrats to the Hex team!
It has been so fun to build this with the team, and it's even more amazing to see all the love we're getting from customers so far! The Notebook agent has really changed the way we're doing our work internally – it's so much easier to get from question to analysis in just a few steps.
If you try out the agent, do let us know what it unlocks for you :) We love to hear your feedback!
At Hex, we've been using the Notebook Agent for a while now so I've been able to use the agent to accelerate one of the toughest parts of product work: pricing analysis.
Normally, pulling together the right data across plans, features, and customer behavior takes weeks of back-and-forth with the data team or a highly paid consultant. This time, I opened a Notebook, asked the agent my questions in plain language, and it instantly wrote the SQL, built the joins, and surfaced charts that actually made sense. What struck me most was the transparency. I could see every query it generated, tweak it when I wanted, and build on top of the results. That combination of speed + trust made it possible to get from “open question” to “real insights” in a couple hours. Even better, when I shared the analysis with our data team, they had no notes. That’s when I realized how game-changing this tool is: it lets me focus on the product decisions, not the mechanics of data wrangling.
The Notebook Agent has already shaped how we think about packaging at Hex and I can’t wait to see what else it can unlock.
Agentic tools have transformed how I design at Hex, allowing me to build and polish my designs quicker than ever.
With the Notebook Agent, I get that same seamless access for data analysis. I can pull the data I need instantly, turning it into real insights that inform and ground my designs. Work that once took hours now begins with a simple question—and quickly becomes a complete analysis in minutes.
We designed the agent to feel natural within the workflows you already know and love in Hex projects. One of my favorite aspects is how transparent it is—you can dive in, see exactly what the agent is doing, and then approve or iterate directly on the cells. It’s powerful, but also intuitive.
Before Product Hunt, I started out my career as a data scientist at a start-up (@aviate ), and then later at Yahoo. I developed some pretty strong opinions about data at an early stage.
In particular, contrary to the "democratize data" movement, I don't believe it's important to make data access easy - instead, I think it's important to make getting RIGHT answers as easy as possible, while guarding against enabling fast, wrong answers. Easy, wrong answers can be toxic to an organization.
Hex is changing my mind on this. It makes it easy for someone to get data in a way that keeps them close to the actual data and schema - by helping them work indirectly with SQL. Before trying Hex Agent, I was a bit skeptical - I wanted it for personal use, but was worried it would be dangerous for non-expert users. After trying it, I think it's certainly a bit risky - but worth the risk. Plus they're closing the gap soon.
In a way, Hex agent reminds me of myself as a junior data scientist - a little green, but up for anything. It's actually made working with data fun again, despite my rusty SQL.
Having used the Notebook Agent since its inception at Hex and after seeing all of the great iterations the team has been making to it, makes me so excited for the world to finally see it, use it, and love it like we do every day at Hex. We are just getting started!!
I have been using the notebook agent as an early adopter and and it's genuinely transformative. I've been using it to build analytics views and products faster than ever. What used to take me a full afternoon of SQL wrestling and chart tweaking now happens in minutes. The agent doesn't just generate code it understands the story I'm trying to tell with data. The magic isn't just in the automation, it's in how it amplifies human creativity. I spend less time on tedious data prep and more time on the insights that actually add value.
Data teams everywhere are drowning in requests while trying to do deeper, more strategic work. Hex's Notebook Agent is the first tool I've used that actually solves this problem and allows me to focus on what the insights mean and not how to build them.
This makes sense. As gen ai development continues to evolve and mature, we're all going to be data scientists.
We absolutely love @Hex at Product Hunt. We moved off Periscope a couple years ago and were immediately blown away by Hex. But, until recently, I felt a big difference between modern AI IDEs like @Cursor and Hex. I have come to rely heavily on Cursor's AI and agent features. It's amazing to see Hex close this gap. My view of the future of technical work is it will be a combination of asking agents high level tasks, fast AI function and syntax autocomplete, and great support for manual code crafting and editing. Love seeing Hex excel in all of these now.
I've been lucky enough to use the NB agent for a while now and it has 100% transformed the way I do data and ML work, and brought it up to the level of AI-enabled efficiency that I get from using claude code etc. for software eng tasks.
I unironically probably could not have built the notebook agent without the notebook agent! So psyched it's out there
Really impressed by how Hex bridges data teams, business users, and AI agents all in one place.
How do you see AI agents evolving on your platform in the next year? Wishing you a fantastic launch!
I've never been more excited about a launch from Hex because the Notebook Agent has fundamentally changed the way data-savvy folks at our company get insights from data. Using the agent, I find myself asking more questions, experimenting and exploring more in areas where I previously didn't have the bandwidth to be curious, and making more headway on my own analyses before looping in our data team for review. Truly an efficiency multiplier (dare I say 10x?). Really proud of the team for building this amazing product, and can't wait to see what others do with it!
Hey Product Hunt! I’m Olivia, a product manager at Hex. Today, I’m excited to introduce the Notebook Agent.
For the last few months, data people have been watching as software engineers have been getting first-class agentic tools to do their work, stuck copying and pasting trying to keep up. We built the Notebook Agent to change that.
Now you can get the best of agentic capabilities in Hex. Just ask a question, get to a first draft in minutes, and iterate on any complex analysis with full control over the agent’s edits.
Why Hex? Because the agent lives right where you work — connected to your projects, files, and warehouse tables. It has the context and tools to understand your analysis and can generate SQL, Python, native visualizations, and Markdown summaries.
We’re already seeing folks use it to get started faster with ad hoc exploration and supercharge deep-dive analysis. If you try it out, let us know what you think!
🖖,
Olivia