AI for your role

AI for Analytics Engineers

Let AI scaffold the models so you can obsess over the tests that make data trustworthy.

Get the Analytics Engineer brief
The shift

How AI is changing the Analytics Engineer role

In 2026, AI copilots scaffold dbt models, write SQL transformations, and draft tests and documentation faster than any Analytics Engineer could by hand. That makes the role more valuable, not less: value concentrates in modeling data so downstream teams can trust it, designing tests that catch silent breakage, and shaping the semantic layer the whole company reasons on.

What AI can take off your plate

  • Writing dbt models and SQL transformations from a spec
  • Generating schema tests and data-quality assertions
  • Drafting model documentation and lineage descriptions
  • Refactoring repetitive transformation logic
  • Explaining a failing test or a data-quality incident

What stays distinctly human

  • Modeling data so downstream trust is earned, not assumed
  • Designing tests that catch silent data breakage
  • Owning metric definitions in the semantic layer
  • Making grain and dimensional-model decisions
  • Balancing warehouse cost, freshness, and performance
Tools

Five AI tools for Analytics Engineers

dbt (with dbt Copilot)
Generates models, tests, and documentation from natural language and centralizes metrics in the semantic layer.
Try it →
ChatGPT
Writes and refactors SQL, drafts dbt tests, and explains a gnarly window function or a failing assertion.
Try it →
Claude
Reviews a transformation for grain and join bugs, proposes edge-case tests, and drafts clear model docs from your code.
Try it →
Cursor
An AI-native editor that understands your dbt project and helps write, refactor, and debug models across files.
Try it →
Datafold
AI-assisted data-diffing and column-level lineage that shows exactly what a change to a model will break downstream.
Try it →
Prompts

Five prompts to try today

Paste these into Claude or ChatGPT and replace the bracketed parts with your own details.

1. Draft a dbt model
I need a dbt model for [purpose] from source tables [describe]. Write the SQL following [conventions], set the right grain, and suggest the schema tests I should add.
2. Design the tests
Here's a dbt model: [paste]. What silent failures could slip through — null spikes, grain changes, broken joins, duplicates? Write dbt tests to catch each.
3. Debug a data-quality incident
A downstream metric jumped [describe]. Here are the relevant models [paste]. Walk through the likely causes upstream and how to isolate which transformation introduced it.
4. Optimize warehouse cost
This model is expensive [paste SQL and describe run frequency]. Suggest incremental strategies, materialization changes, or refactors to cut cost without hurting freshness.
5. Define a semantic metric
We want a governed definition of [metric] in the semantic layer. Propose the definition, the dimensions it should be sliceable by, and the tests that keep it honest.
The playbook

Every AI play for Analytics Engineers

The full library of tools, prompts, and tricks for your role — updated every week. Tap any card for a step-by-step walkthrough and examples.

✦  New tools, prompts, and tricks are added every week — and go straight to subscribers in their morning brief. Skip the scrolling and get yours delivered free. Get my free brief →
Loading the library…

A day in your inbox

This is the kind of brief a Analytics Engineer gets, every weekday morning.
Monday morning
✦ Personalized for: Analytics Engineer
Today's Tool
Generate the model, own the tests
Let dbt Copilot scaffold the SQL, then spend your judgment on the tests and contracts that catch silent breakage. Transformations are easy to write and easy to get subtly wrong.
Today's Prompt
Ask what could silently break
Paste a model and ask AI for the failures that wouldn't throw an error — grain changes, dupes, null spikes — then codify each as a test. That's where trust is built.
Today's Trick
Put warehouse spend on an AI audit
Have AI scan your query logs for the most expensive models and propose incremental or materialization fixes. Cost control is becoming a headline skill.

Get the Analytics Engineer brief

One AI tool, one prompt, and one trick for Analytics Engineers, every weekday morning. Free.

Almost there — we just emailed you a confirmation link. Click it to activate your brief.
Free forever. Unsubscribe anytime. We use your role only to personalize your brief.