Matter Labs

Matter Labs needed insights into the activity on their network, team health metrics, and economic prediction models for proposed features, however their existing stack was not set up to reliably create and share these types of insights.

Challenge

Matter Labs was using a combination of query editors, CSVs, and Google Sheets to do analysis which were unreliable and difficult to maintain. They were using custom scripts to populate their warehouse BigQuery. These custom pipelines would often break and it was time consuming to fix them to keep providing teams with the data they needed.

They needed to invest in infrastructure and tooling that would help them produce reliable, on-time, professional looking reports.

Solution

Their data team of one, Landon, decided to set up a data stack consisting of Mage for data pipelines, keeping Google BigQuery for their data warehouse, and using Hex for both dashboarding and ad hoc analysis. Landon chose these tools because they allowed him to get started quickly, worked for a variety of use cases, and could scale with their business.

Mage is an open-source data pipeline tool that allows you to use Python, SQL, R, and PySpark to transform data. This flexibility made it simple for Landon to connect several disparate data sources to BigQuery quickly without a lot of maintenance overhead and cost savings (since it’s open-source). With the data loaded from HubSpot and multiple 3rd party APIs, Landon chose Hex to explore and share insights.

Landon chose Hex because it allowed him to both create company wide interactive reports and in depth ad hoc analyses. He also loved the ability to combine SQL, Python, and no-code in one notebook so he could do all his querying, analysis, and visualization in one place. He was able to complete work quicker, in a reproducible way, and make it automatically available to everybody with access.

With Mage, BigQuery, and Hex he streamlined the process of delivering analytics at Matter Labs

Results

By using Mage and Hex, the time to go from doing getting started tutorials to production ready data pipelines and dashboards was less than three days each, compared to what would have been several weeks for competitors.

With this data stack Landon can now build interactive reports to track general health metrics for every team and monitor how the network is performing easily as a one man data team in a few hours. Hex’s notebook style interface and app builder has made it easy for him to produce reliable, on-time, professional looking reports.