off-the-stack
cd ~/careers
Modern dataaka "AE"

Analytics Engineer

The unglamorous data role that quietly out-earns and out-hires the data scientist everyone wanted to be.

Entry
$90k
Mid
$130k
Senior
$175k+
Demand
Rising

Analytics engineering sits between data engineering and analysis: you build the clean, tested, documented data models (mostly in SQL and dbt) that the whole company trusts to make decisions. It emerged from the modern data stack, it's in heavy demand, and it's far less crowded than the data-science track that soaked up all the hype.

The myth

It's just writing SQL queries for dashboards.

The reality

You're a software engineer for data: version-controlled, tested, documented data models with CI, treating analytics like a real codebase rather than a pile of ad-hoc queries.

cat ./what_you_actually_do.md

  • Build and maintain transformation models in dbt that turn raw data into trusted, documented tables.
  • Apply software engineering to analytics: version control, testing, CI, and code review on data models.
  • Define the metrics layer so 'revenue' means one thing across the whole company.
  • Partner with analysts and stakeholders to turn fuzzy questions into reliable data products.
  • Own data quality and documentation so people actually trust the numbers.

cat ./why_underrated.md

Everyone chased 'data scientist' for a decade, leaving the role that actually makes data usable comparatively empty. Analytics engineering is newer, less hyped, and therefore far less competitive to enter — while being more directly valuable to most companies, which need trustworthy reporting far more than they need another ML model. The skill set is concrete and learnable (SQL is the core), the tooling is approachable, and the work is the backbone every data team quietly depends on. It's the high-leverage data job nobody told you to want.

grep -i 'good fit' ./who.md

  • People who like bringing order and trust to messy data.
  • Pragmatists drawn to impact over the prestige of 'AI'.
  • Detail-oriented builders who enjoy SQL and clean modeling.

cat ./pay.md

Demand outstrips supply because the role is newer than the data-science gold rush and far less saturated. Senior analytics engineers and analytics leads reach $160k+, and the skills (SQL, dbt, warehouse modeling) transfer cleanly across nearly every company that has data — which is all of them.

./break_in.sh

  1. Master SQL, then learn dbt

    SQL is 80% of the job; dbt is the tool that turns it into engineering. Both are free to practice and well documented.

  2. Build a portfolio project

    Take a public dataset, model it properly in dbt with tests and docs, and put it on GitHub. That's a hireable artifact.

  3. Learn one cloud warehouse

    Snowflake, BigQuery, or Databricks. Free tiers let you build real models on real-ish scale.

  4. Pivot from an adjacent seat

    Data analysts and BI folks are perfectly placed to level up into analytics engineering — it's one of the most accessible on-ramps in data.

tail -f ./a_day.log

  • 09:00A stakeholder flags a number that looks wrong; trace it through the dbt lineage to a bad join.
  • 11:00Refactor a sprawling model into clean, tested, documented layers.
  • 14:00Define a new metric in the semantic layer so every dashboard agrees on it.
  • 16:00Review a teammate's data-model PR like the real code it is.

ls ./toolbelt

  • SQL
  • dbt
  • Snowflake / BigQuery
  • Git
  • A BI tool
  • Python (light)