Trace, Debug, Deliver: June Updates from Recce

It’s been a pretty busy couple of months with conferences and some exciting new features from Recce! Our new column-level lineage feature helps teams focus on what matters most when doing review and our new recce debug command helps address common problems.

Here’s what we’ve been busy working on:

New feature: Trace impact through with column-level lineage

Column-level lineage enables you to trace data as it is transformed through your data project. At each stage you can see which column uses that data and how.

Click any column to view:

• The impact-radius for a column: Downstream models and column dependencies.
• The transformation type: How the column is used.

When used together with breaking-change analysis (which assesses the likelihood of a SQL change to impact data) column-level lineage provides powerful and actionable insight for the following tasks:

• Exploration during development.
• Assessing data impact following model changes.
• Root cause analysis when debugging issues.

Checks the docs on how to get started with column-level lineage.

New command: Recce debug

If you’re having trouble with Recce, then the recce debug command can help you understand why. The new commands tests the following:

• dbt artifacts exist for your development environment in the `/target` directory.
• dbt artifacts exist for your base (stable) environment in the `/target-base` directory.
• Your data warehouse connection is functional.

If any of the tests fail, you’ll be given a helpful hint on what your next step should be.

Ensure Recce is ready-to-go with the new recce debug command.

Recent blog posts

In our most recent blog posts we explore why column-level lineage is useful, how data diffs are just one tool in your toolkit, and a better way to build a time spline. Check out more below..

How Recce performs column-level lineage

Check out this deep-dive into how Recce constructs column-level lineage by analyzing columns to understand references, dependencies, and also classify transformations. Read more.

More than just a diff

Data diffs are becoming more common, but not all data teams are using them wisely. In this article we explain why data diffing is just a tool, and shouldn’t be the goal. It’s the underlying workflow that Recce enables where the real value is. Read more.

Recce IRL

This month members of our team attended the Databricks Data + AI Summit, spoke at San Francisco’s ML Ops Community June meet-up, and helped organize Taipei’s dbt meet-up. We loved getting to meet new faces and seeing existing Recce community members. We hope to catch you at an IRL event soon. Follow our official social channels for more.

Below is a picture of our Product Karen & Head of Data Dori at ODSC East.

We’re always looking to learn from real teams using Recce!

Whether it’s feedback, setup tips, or roadmap talk, we’d love to connect, you can book time with us here.

Keep shipping,

The Recce Team