You Asked. We Listened.

Dataiku Product Catie Grasso, Christina Hsiao

How often is it that you leave a suggestion for how to improve a product you love and the suggestion actually gets implemented? The answer is probably very rarely, if ever. You might even think to yourself, “I wonder if anyone will even read this comment or if this form is simply to make it look like the company listens to customer feedback?” Sigh. 

Well, call us unconventional, but at Dataiku, we consider customer suggestions one of our key inputs to product direction and roadmap. Not only do we have a dedicated listening post to collect your ideas, but our product managers and software engineers personally review the suggestions and keep the community apprised of any product developments stemming from a submission. We’re thrilled to share that in our latest release, Dataiku 11, we delivered solutions for not one, but six Product Ideas!

That's not all, though. When we also include requests and bug reports sourced from our customer support teams and technical support interactions, it's much more than six. In fact, 400 requests were delivered during the eight-month period leading up to and culminating with Dataiku 11. Wow!

Read on to see which new features were influenced by customer suggestions and learn more about how you can submit your innovative ideas in the future.

What Is the Product Ideas Board, Anyway?

Launched last year, the Product Ideas board is an interactive space where members of the Dataiku Community can submit suggestions to help us improve the platform and keep track of our responses. Dataiku product managers and developers regularly review the ideas and provide immediate solutions or status updates as feature requests move through the review process. 

Product Ideas Board

In addition to your own submissions, you can get involved by reviewing, following, and commenting on other users’ suggestions. And don’t forget to upvote the ideas you like to add weight to these requests!

6 Product Ideas Delivered in Dataiku 11

One theme that surfaced multiple times in Product Ideas was that our most technical users would be happier and more efficient without needing to use remote IDE integrations or to copy/paste code into their favorite editor in order to access productivity helpers like automatic code completion, built-in function documentation, multi-line commenting, markdown, and interactive debugging. The Product ideas highlighted in the image below are support for JupyterLab, replacing the code editor, and multi-line commenting in code recipes.  

3 distinct Product Ideas related to the theme of productivity tooling and free choice of code editors: support for JupyterLab, replacing the code editor, and multi-line commenting in code recipes

To meet this need, Dataiku delivered Code Studios — personal development environments connected to a Dataiku project that run popular web-based IDEs such as VS Code, RStudio, and JupyterLab. Code studios enable developers to enjoy a familiar coding experience when building recipes or web apps, all in a fully secure environment powered by an elastic Kubernetes backend. 

Status updates on a submitted product idea from a Senior Product Manager and VP of Engineering

Giuseppe, one of Dataiku’s Neurons (our global network of data leaders and practitioners who make exceptional contributions to the Dataiku Community) suggested that adding some Feature Store functionalities such as a centralized, governed feature registry and discovery function would save teams lot of time by preventing duplication of code and efforts. Marlan, another Neuron who had previously published insightful Community content on how his organization had built a feature store using Dataiku, chimed in to the Product Idea thread to provide additional context based on his experiences. 

As this request aligned with our broader product principles of collaboration and reuse, we are happy to report that roughly six months later we delivered a solution with the release of Dataiku 11. The central feature store enables teams to discover, explore, and reuse reference datasets containing curated features, and reduces the time data teams spend reinventing the wheel.

Status update on this Product Idea from Dataiku's VP of Engineering

Brand new features like these are certainly exciting, but feedback from real-world data practitioners is also critical to helping us improve the user experience for more mundane daily tasks, such as routine data wrangling and visualization. 

It'd be great if the visual filter builder were able to support more sophisticated logic. Currently, all the filtered values need to be combined with either and or or, but in some cases, users might want to use nested logic without needing to write an expression using formula language or SQL. This UI handles nested logic pretty well. It'd be awesome if logic like this could be expressed visually, I think it'd speed up development quite a bit and also help users intuitively understand what they're filtering for." - @natejgardner, author of the Product Idea

The idea above led to no less than four different enhancements to the way users construct and manage project logic. New preparation processors, a switch formula, and a redesigned visual framework for constructing if...then...else expressions throughout the entire user interface make it simpler for teams to recode data values and embed conditional logic as business rules in projects.

Finally, building charts and dashboards to visualize data is an important part of Everyday AI. A user request for an easier way to display categorical aggregations in charts led to a quick win for Dataiku’s product team and a more satisfied customer. Now, all aggregated charts (columns, bars, pies, lines, areas, pivot table, etc.) support the “Count Distinct” and “Count Non-Null” aggregation functions, which as a bonus also makes it possible to have non-numerical measures.

User requested feature: display distinct category counts as an aggregated chart measure

It's More Than Lip Service

Some companies might have front line teams like customer service or social media managers who are responsible for reading product ideas and recommendations solely to diffuse risky situations with disgruntled customers — and the comments don’t ever actually get surfaced to product managers. At Dataiku, we don’t just listen to de-escalate issues, but rather to learn and improve. We always want our users to know that we hear them and value their feedback to help inform our future roadmap and vision. So, if you want to get involved and be a citizen product manager of Dataiku, follow the steps below to share your Product Ideas. We want to hear from you!

You May Also Like

Accelerate Financial Forecasting With Dataiku

Read More

Keep AI Under Control With Dataiku 12

Read More

Build Better Customer Relationships With Next Best Offer (NBO) for Banking

Read More

Making GPT More Accessible With Dataiku

Read More