All Articles
AirOps Updates

dbt + AirOps AI Data Sidekick

Write, optimize, and troubleshoot dbt SQL queries. Build Python models. Faster than ever before.

Published on Jan 17, 2024 by Kyle Dempsey

We’re thrilled to share a new tool that will help dbt users write, optimize, and troubleshoot their SQL queries faster than ever before: AirOps AI Data Sidekick.

Data Sidekick is free, and includes a growing library of AI-powered recipes that help data practitioners do more in less time. It’s available anywhere you work, including inside of the dbt Cloud IDE, dbt Core, Snowflake, AWS QuickSight, Mode Analytics, and more.

You can learn more about Data Sidekick here, or keep reading to see exactly how Sidekick makes you more efficient inside dbt Cloud 👇

Data Sidekick recipes for dbt users

As a dbt user, Sidekick lets you instantly generate…

  • dbt SQL with Jinja
  • dbt YAML files for your models
  • Full Python scripts that can run directly in dbt

Let’s take a closer look at each one. You can also check out a full demo from our very own Kyle Dempsey here.

1. Use natural language to automatically generate a dbt SQL query

Sidekick’s dbt SQL Writer recipe lets you input any natural language request, and with that prompt, Data Sidekick will generate a full SQL query that references multiple dbt models using Jinja.

Since you’re already working in dbt Cloud, you can also use all of the available tools, like running the query and building your model.

2. Use Data Sidekick to automate dbt model documentation

Once the query is written, it’s time for everyone’s favorite step: Writing the YAML file entry for the model that you just created.

Sidekick makes that step quick and easy, too – paste the query directly in Sidekick using the Generate dbt Config recipe, and boom, instant YAML entry. ⚡Once you paste that into dbt, you’ll notice that Sidekick also generates the name and definition for each field and automatically adds tests.

The entire process takes mere seconds, so now there’s no excuse not to do it!

3. Generate complete Python scripts using the correct dbt syntax

Data Sidekick can also generate powerful Python scripts!

Create a new Python file in your dbt Cloud environment, then use the dbt Python Model Creator in Sidekick to generate a complete Python script that uses the correct dbt syntax. 

For example, prompt Sidekick to use the Prophet library, then ask it to “Forecast cost into the future for the next 6 months.”

From there, you can choose the model that you just created with SQL, generate a Python script that inputs that model as a data frame, and paste it directly into your dbt Cloud environment. From there, Sidekick will return the final data frame as a forecast to be stored as a table back in your data warehouse.

Work with data 10x faster, anywhere

AirOps AI Data Sidekick can help anyone who works with data move faster and focus on the most important, complicated tasks. 

In addition to these dbt-focused recipes, you can also use Sidekick to: 

  • Suggest questions you can answer from any database table 🧐
  • Explain, modify and fix existing SQL queries 🔧
  • Optimize queries to improve performance and trim your warehouse bill 💸

Want to give it a try? Sidekick is free for personal use and small teams. 🎉

Click here to get started

And if you have ideas for additional recipes that would make your job faster and easier, let AirOps know! We’re always interested in hearing from our fellow dbt community members – feel free to reach out to me directly on LinkedIn to share your first impressions, feedback, and ideas!

Work with data 10x faster.

Magically draft, correct, and explain SQL. Instantly write Python scripts. Free for individuals and small teams.