Blogs in this series are based on the early access preview version of Plotly Studio. Plotly Studio is an AI-powered desktop application by Plotly designed to automate the creation of professional data apps and visualizations.
Exploring Plotly Studio: A Promising Start with Data-Driven Dash Apps
Over the past few weeks, I’ve seen several impressive applications built with Plotly Studio shared across LinkedIn. Inspired by these examples, I decided to explore the tool myself—starting with something simple yet practical.
I uploaded a familiar dataset and asked Claude to generate an initial outline (or prompt) for the app. Within minutes, I had a working prototype in preview mode that allowed me to interact with the data through filters and visualizations. I approached the experience as someone new to Dash and Plotly, evaluating the usability and accuracy from a fresh perspective.
About the Dataset
The dataset I used contains financial data—specifically monthly and yearly revenue and freight amounts—broken down by product and product category. In other words, a classic scenario for data-driven exploration and dashboarding.
First Impressions
The preview app was functional and surprisingly insightful right from the start. Filters worked as expected, interactions were smooth, and most importantly, the data in the charts appeared accurate.
There was one exception: the freight data turned out to be misleading. The issue wasn’t with the app but with the input itself—garbage in, garbage out. The metric used was technically correct, but the data wasn’t. This highlighted a broader point: if there’s no visible error and the output still looks plausible, how can we reliably detect something’s off?
Interestingly, if Plotly Studio were to support uploading multiple .csv files, it would eliminate the need to pre-join the data externally—saving users from potential join errors. That’s a feature I’d definitely welcome.
Reflections on Filters and Usability
The app included a wide range of filters—some in a dedicated filter panel, others embedded directly in the charts. While this flexibility is powerful, it did cause some confusion. For example, applying a filter in one area didn’t always reflect in others, leading me to wonder whether the filter had been applied at all.
Of course, this is partly on me: I had requested a broad set of filters to begin with. Still, from a UX perspective, it would be helpful if the scope and behavior of filters were more consistently communicated.
Final Thoughts
Overall, my experience with Plotly Studio was very positive. For an early-access tool, it’s already quite robust and shows tremendous potential. I can only imagine how much time and effort this would have taken to build manually.
I’ve summarized my approach, results, and initial observations in a short PDF (see below). If you’re curious about Plotly Studio or considering it for your own data projects, I highly recommend giving it a try.
test_plotly_studio