We use analytics cookies.We use cookies to analyze site traffic and improve your experience.
Loading...
System Design Masterclass · Data Governance Compliance
Free preview: the first 2 parts of this 9-part lesson.
Already bought the course? Sign in to read the whole lesson.
this lesson is part of
System Design Masterclass
Unlock this lesson and the 751 other paid lessons
$25 once, by PayPal
See the System Design courseread the rest of this lesson in
System Design Masterclass
Unlock this lesson and the 751 other paid lessons
$25 once, by PayPal
See the System Design coursein Data Governance Compliance
This lesson is part of the full System Design Masterclass. Unlock it along with 769 other lessons for a one-time payment.
One-time payment by card or PayPal
Get Lifetime AccessOne-time payment: UPI, card or netbanking
Get Lifetime Access18 Foundation lessons are free, no payment needed. Try them now
It is easy and good
“It is really nice and good Please join the course”
A mid-size company has 200 employees, 15 databases, 40 , and 8 third-party integrations. The marketing team creates a customer segmentation using data from the CRM. The product team builds a recommendation engine using data from the analytics warehouse. The finance team generates revenue reports from the billing database.
All three teams use "customer data." But they define "customer" differently. Marketing includes leads. Product includes free trial users. Finance includes only paying customers. Their reports contradict each other. The CEO asks "how many customers do we have?" and gets three different numbers.
Nobody knows which number is right because nobody owns the definition of "customer." Nobody owns the data quality. Nobody set standards for how customer data should be collected, validated, or reported. There's no data governance.
LinkedIn ran into exactly this problem at scale. With hundreds of data pipelines and thousands of datasets, they couldn't answer basic questions about data freshness, accuracy, or ownership. Their response was building DataHub, an open-source metadata platform that became the backbone of their data governance. DataHub now tracks 500,000+ datasets at LinkedIn and has been adopted by companies like Acryl Data, Optum, and Saxo Bank.
Data governance is the organizational framework that establishes who is responsible for data, what standards apply, how data quality is maintained, and how data decisions are made. It's not a technology problem. It's a people and process problem with technology enablers.
| Component | What It Defines | Who Is Responsible |
|---|---|---|
| Data ownership | Who is accountable for each data domain | Data owners (usually business leaders: VP of Sales owns customer data) |
| Data stewardship | Who maintains data quality day-to-day | Data stewards (usually senior analysts or data engineers) |
| Data catalog | What data exists, where, what it means | Platform/data engineering team, with stewards populating descriptions |
| Data dictionary | Definitions of every term and metric | Data stewards + business stakeholders in the governance council |
| Data policies | Rules for collection, use, sharing, retention | Governance council (cross-functional) |
| Data quality standards | Acceptable quality thresholds per dataset | Data stewards + data owners (e.g., "email validity rate must be > 95%") |
| Access control policies | Who can access what data, under what conditions | Security + governance council |
| Metadata management | Tags, lineage, classification for all data assets | Platform team + automated tools (dbt, Airflow, catalog integrations) |
The Data Governance Council: A cross-functional group (engineering, product, legal, finance, security) that makes data-related decisions. "Should we share customer data with this partner? What's our policy on data retention for GDPR? How do we standardize the definition of 'active user' across all teams?" The council meets monthly, publishes decisions, and resolves disputes. Without it, every team makes independent data decisions that eventually contradict each other.
Spotify's data governance council includes representatives from every major business unit. When a new ML model needs access to user listening data, the council evaluates the request against their data usage policies. This prevents the "move fast and ask forgiveness" pattern that leads to privacy incidents.