Noteable: The Interactive Notebook Document for Modern Data Teams
In an increasingly data driven world, we’re all seeing a continuing surge of investment, tooling, and advancement in how we create experiences with our data. There no longer exists a stark divide between research groups or technology companies and more traditional industries. Computers and devices drive our businesses directly or indirectly, both taking and generating a wealth of data in the process. As a result, we each have data that needs to be understood and turned into useful information to compete, influence, or succeed.
However, the interfaces and tooling which enable access to information still lag far behind the average organization’s needs. This friction is driving both the record-breaking number of launches and the size of IPOs in the data space. Despite this growth, there’s still a wide gulf separating expected ease of use and actual complexities when working with data. Even in leading technology companies, specialists and analysts are rendered less effective due to inadequate tooling to complete their job where it intersects with data; meanwhile, less technical-focused users are left behind entirely. While at Netflix, we even gathered these companies — Amazon, Microsoft, and Bridgewater, to name a few — at our Notebook Summit, and the stories we heard from companies large and small repeatedly echoed this common problem.
Enter Noteable. For the past year we’ve been building a remarkable executive team with deep knowledge of these challenges in the data domain: Michelle Ufford, Elijah Meeks, Matthew Seal, and Pierre Brunelle. Each of us have worked extensively on a range of data problems throughout our careers and, notably, have utilized a common technology which was helping unify data usage at Netflix: Jupyter Notebooks. Integrating this technology into Netflix’s data platform enabled teams with no direct engineering support to solve real business needs and share what they were doing with their colleagues. This sparked a wave of efforts in other companies to build similar tooling and coordinate in the open-source spaces between.
“There is a happy world of dynamic and interactive analytics waiting for you!”
— Semi Koen: Jupyter is the new Excel
The investment in Notebooks had extreme returns for the effort, and allowed very small teams to help large organizations succeed in ways which those same investments elsewhere could not have achieved. And this model was not isolated to Netflix. You can point a finger at just about any large technology company and you’ll find at least one Notebook team and a similar story arc emerging or already present. But the value of and demand for easy data manipulation interfaces is not isolated to companies who can invest half a dozen or more developers to enhance these technologies to meet their business needs. They are in demand everywhere, by virtually every organization you know about, from non-profits, to schools, to entertainment centers, to manufacturing companies, and even restaurants.
We believe that the ability to make better, data-informed decisions should not be isolated to the few organizations able to spend the money to build the necessary tooling advances for themselves. Driven by this belief, and supported by the clear needs of the organizations we spoke with, we decided to launch a new company. After founding Noteable in May of 2020, we began to develop the first enterprise-grade notebook platform for data-driven teams. We want to both enable new use cases and be friends of Jupyter, supporting the open source standards and promoting the community therein. We’re energized to be solving these problems and working towards our central goal: allow people to work with their data the way they want.®
If you’re not familiar with Jupyter Notebooks yet, they’re traditionally a Data Science tool for exploring data and solutions against that data incrementally. They provide an interface for defining small cells of code or documentation in an ordered list that you can rerun and adjust until they each solve the next step of a larger problem. Each cell supports rich outputs including visualizations, tables, formatted text, and any other media type you can imagine. This format makes guided data interactions easy to follow, share, and ultimately rerun.
By sitting between these worlds, notebooks combine the best attributes of BI, visualization, and coding interfaces while lowering the entry point for new users.
Notebooks differ from BI tools and visualization systems in that they provide embedded code snippets that can express the full range of high level programming languages while still allowing for visual reflections of the data retrieved by this code. They differ from traditional coding environments in that they better encode documentation, learnings, and examples in a single shareable document. By sitting between these worlds, notebooks combine the best attributes of BI, visualization, and coding interfaces while lowering the entry point for new users.
What we’ve discovered over the past half decade is that beyond the traditional data science use cases, Notebooks achieve a guided experience that is even more valuable to data users who have fewer tools available. A hidden cost to many teams today is onboarding and maintaining a dozen different interfaces to perform analysis on one’s data. We’ve seen firsthand the value in a common shared interface that can democratize technologies and bring them to a much broader audience, giving more users across an organization access to deeper, actionable insights. It’s a literal workforce multiplier.
But at the same time, we want to build the solutions for problems outside of the scope of those [open-source] efforts and provide a best-in-class experience
The open-source offerings for libraries and hosted services in this technology space have both gained a lot of traction and also advanced significantly in the past few years. One of the larger efforts to provide a usable service in this space has been JupyterHub. It gives auth and containerization with a hosted Jupyter interface that can get an organization off the ground.
But for teams becoming ever more data-driven, the need for collaboration, organization, and better data integration goes beyond what’s available today. Noteable aims to tackle these concerns and enable an exciting data experience extending the groundbreaking work done in the open-source.
To do so, we started with and continue to recruit engineers who actively contribute to and organize efforts within the open-source community. These excellent people also support and build many related projects in the open-source space including: Papermill, Semiotic, Data Explorer, Nteract, and Python. We’ll continue to invest our time and resources into improving Jupyter and its neighboring libraries. But at the same time, we want to build the solutions for problems beyond the scope of those efforts and provide a best-in-class experience.
To reach these goals, we’re also looking into solving real user stories beyond the ones we know best. This effort has led us to collaborate with a range of amazing design partners who have shared their needs, skills, and knowledge to refine our ideas and implementations. We’re thrilled by these relationships and the engagement in the process so far. Without these interactions we’d build what we know instead of what’s really needed by our customers.
With our initial major milestones coming to completion this summer, we’re ever more focused on providing this next generation Notebook experience. We have the right people, the right talent, and the funding to allow us to achieve our vision. If you want to join us on this journey or become a beta user, sign up for our beta email list.
If you want to read more about why developing your data culture is key to the success of your organization, take a look at Noteable CEO Michelle Ufford’s The Leader’s Guide to Being Data-Driven in 2021 (Part 1).
Interested in learning about our approach to data visualization in our product? Noteable Chief Visualization Officer Elijah Meeks shares his thoughts on Designing for the Data Visualization Lifecycle.
