Join us on April 7-9, 2021

The first industry event focused on data engineering

Register Today
All Blog Posts

Visualizing Multi-Dataset Wrangling with Flows

May 25, 2017

In the latest release of Trifacta, we’ve introduced a new interface for managing relationships across your datasets. Say hello to Flow View. You can now visualize how your complex wrangling projects incorporate multiple datasets, experiment with exploratory transformations, and reuse past work to save time. To realize these benefits, we’ve designed project spaces called Flows” that illustrate the relationships of your crafted transformation logic Recipes” which link your input and resulting data objects, respectively Imported Datasets” and Wrangled Datasets”, together. With this foundation, we envision supporting a broader range of use cases requiring more complex sharing and operationalization in the future. 

See many things clearly
With Flow View, users can efficiently navigate their complex data wrangling tasks. People often combine, reference, and compare up to dozens of datasets early in the data preparation stages. Rather than requiring our analysts to maintain a complicated mental map or draw out diagrams manually for every join and union, we anticipate reduced cognitive load by showing this higher-level view automatically. A single view clearly illustrates all of the Recipes and how they leverage Datasets and more importantly the interdependent relationships between them.

Experiment with confidence
With copies, experiment with confidence. If you want to test a variation of a transformation Recipe on the same data, have no fear; simply make a separate copy of a Recipe to branch your work without affecting the original Recipe. You can compartmentalize several explorations, then once satisfied with one, you have the option to delete nonessential explorations. Alternatively, make a copy of an entire Flow to support explorations at a larger scale.

Reuse, your time is valuable
Flows, Recipes, and Datasets can be reused and remixed as templates to save you time rather than starting your work from scratch. Instead of importing new, similar data and laboriously rebuilding entire Flows every month, copy a Flow and simply replace the Imported Datasets that feed the Flow although some minor tweaking may be required. Also, start a new Flow with a reference to already transformed data via Wrangled Datasets” from other Flows to help compartmentalize commonly used subtasks. 


To learn more about the latest features in Flow View, check out our blog on the Spring ’17 Release and sign up for Trifacta Wrangler to try it out yourself.

Sign up for Free Wrangler

Looking ahead
As always, tackling design challenges like these align with Trifacta’s deep commitment to crafting intuitive data-preparation experiences for non-technical analysts everywhere. While Flow View immediately supports basic use cases, we envision this effort as groundwork for more ambitious initiatives in delivering a collaborative, cloud-optimized, enterprise solution. This new paradigm will pave the way for team members to prepare data collectively and iteratively before automating into productionizing Flows.