iRobot is the leading global consumer robot company, designing and building robots that empower people to do more both inside and outside of the home.
Conversations around data were stuck on semantics. Data users didn't know all of the data that existed, what it all meant, how to use it and if they could trust it. That made many product and customer insights hard to find and trust. All that tribal knowledge meant even if someone had access, data was only usable by a select few within the company and using such data slowed things down.
One of iRobot's strategic goals is to develop closer relationships with its customers and offer them more personalized cleaning experiences. This has required iRobot to more quickly learn from its customers and leverage these insights to deliver new experiences. To meet this objective, it was vital to provide teams with data products and experiences that help them use data in their day-to-day decision making while trusting that data is fit for purpose and being used within the bounds of iRobot's terms of service and strict privacy and security regulations.
To this end, iRobot decided to invest in a data catalog, so they could democratize access within the company while also ensuring that the right guardrails were in place. Michelle Gulen, manager of Data Analytics, explained that manual documentation and lineage tracking wouldn't have been sustainable. Gulen explained "We needed an automated data catalog and not rely on manual documentation of data that gets out of date as soon as it’s written".
iRobot needed an automated data catalog that'd automatically ingest, infer and surface metadata from its Amazon Athena Data Lake as well their BI tool, Mode.
Michelle Gulen shared "Before you can use the data, you have to know what data you have, what it means and how to use it in the right way".
iRobot chose Amundsen because of its focus on automation and the integrations in the cloud ecosystem. Additionally, Amundsen's large community of users means that the product will innovate fast and its integrations will evolve as the larger data ecosystem changes. After the initial Amundsen deploy, iRobot moved from Amundsen to Stemma due to its ease of use and features that lowered time to value and its speed of delivery.
iRobot’s Stemma deploy includes automated metadata about how often a table is updated and what other tables it is commonly used with. This provides iRobot’s data users with additional context about delivery expectations relative to the data and how to query it without being bogged down by manually documenting that information for all its data.
The data science and analytics teams at iRobot use Stemma almost every day. It's helped these teams get context about data - what data is there, what it means and how to use it - really quickly. All that without the need to document every single detail that gets easily out-of-date. "Data users can go to Stemma to find what they need without having to track down three different subject matter experts to know what's going on", said Gulen.
Stemma has helped to increase the speed at which iRobot's research and development teams are able to make product-related decisions. Instead of being stuck in “do we have this data” mode, they are now able to build, ship and analyze improvements faster than ever before.