Fintel AnalyticsBigQuery Service Account Key in dbt Repo: Find & Fix It
A BigQuery service account key committed to a dbt repo is a live credential, not a tidy-up task. Find it, rotate it and remove it for good.
Fintel Analytics blog
Data & AI Insights
Expert perspectives on data analytics, AI agents, internal tools, and business intelligence from the Fintel Analytics team.
9 of 206 posts
Over-Broad IAM Grants in BigQuery: Find & Fix Them Now
Over-broad BigQuery IAM grants are the most common—and most quietly dangerous—misconfiguration in fast-growing data teams. Here's how to find them in under five minutes.
BigQuery Access Control: Row & Column Security Done Right
Most growth-stage companies running BigQuery give everyone too much access. Here is how to implement row-level and column-level security that scales with your team.
PII Sitting Unmasked in BigQuery: How to Find It Fast
Unmasked PII in BigQuery is a compliance and security risk most teams discover too late. Here's how to find it in under five minutes using SQL against INFORMATION_SCHEMA.
dbt Project Governance for Startups: Stop Model Sprawl
Model sprawl is the silent killer of growing dbt projects. Learn how to govern your dbt project before undocumented models, broken DAGs, and degraded test coverage cost you.
Undocumented dbt Models: Find Every Mart With No Description
Run two commands and one BigQuery query to list every undocumented dbt model in your project, rank them by real usage, and stop new ones shipping.
Dashboard Sprawl Audit: Retire What Nobody Trusts in 2026
Too many dashboards and no single answer to 'what is our revenue?' A practical dashboard sprawl audit: inventory, score, retire, and prevent the problem returning.
Self-Serve Analytics for Business Teams: Build It Right in 2026
Self-serve analytics promises to free your data team from the dashboard backlog — but without the right foundations, it creates more confusion than it solves. Here's how to build it right.
SQL Semantic Layer: Why Your Metrics Are Broken in 2026
If finance and ops see different revenue figures, you don't have a dashboard problem — you have a semantic layer problem. Here's how to fix it.