Proud Ndlovu
dbt · BigQuery · SQL | Star Performer | Open to Remote
New to Contra
Proud is building their profile!
I built a production-style analytics pipeline from scratch this week.
4 external APIs → Python ingestion → BigQuery → dbt staging + marts → automated tests → Looker Studio executive dashboard → GitHub Actions running it daily at 6am UTC. Zero manual intervention.
The domain is crypto. The architecture is the point.
Swap the 4 crypto APIs for Shopify, Stripe, Segment, and Salesforce — the ingestion pattern, staging logic, mart layer, and test suite are identical. This is the same structure that powers data teams at scale.
Two real issues caught during the build:
→ BTC returning NULL from the exchange rate API — caught by dbt tests, fixed at ingestion
→ FX scorecard summing rates instead of averaging, producing USD/ZAR of 163.9 instead of 16.41 — caught by human eyes on the dashboard
That's why you need both automated testing AND visual validation. Neither replaces the other.
Full pipeline on GitHub. Live dashboard linked in the repo.
🔗 github.com/ApostolicDA/coingecko-analytics-pipeline
#AnalyticsEngineering #dbt #BigQuery #Python #DataEngineering #OpenToWork
1
13