Skip to main content

​
What is Knowledge Builder?

Knowledge Builder is SkoutLab’s continuous learning system. Once you connect data, it works in the background to:
  • Understand your data structure and relationships
  • Monitor for changes, anomalies, and trends
  • Build institutional knowledge about your business
Think of it as an analyst who’s always reviewing your data, even when you’re not asking questions.

​
How It Works

When you click Start Learning, Knowledge Builder begins a continuous cycle:
1

Schema Discovery

Scans all tables, columns, and data types. Identifies primary keys, foreign keys, and relationships.
2

Semantic Understanding

Goes beyond schema to understand what fields mean:
  • “revenue” vs “sales” vs “income” — same concept?
  • Status codes: what does “1”, “2”, “3” mean?
  • Currency fields: USD? Local currency?
3

Pattern Recognition

Finds common patterns in your data:
  • Typical value ranges
  • Seasonal trends
  • Normal vs abnormal distributions
4

Continuous Monitoring

Watches for changes and anomalies:
  • New data arriving
  • Unexpected values
  • Schema changes

​
What Knowledge Builder Creates

​
Data Understanding

For each data source, Knowledge Builder documents:
Knowledge TypeDescription
StructureTables, columns, relationships
SemanticsWhat fields actually mean in business terms
QualityData issues, missing values, inconsistencies
PatternsCommon query patterns and best practices

​
Alerts & Insights

Knowledge Builder can detect:
  • Schema changes — New columns, dropped tables
  • Data freshness issues — Data not updating as expected
  • Anomalies — Values outside normal ranges
  • Quality degradation — Increasing null rates, duplicates

​
Starting Knowledge Builder

1

Go to Data Page

Navigate to the Data section in the sidebar.
2

Click Start Learning

Click the Start Learning button in the Knowledge Builder panel.
3

Let It Run

Knowledge Builder runs continuously in the background. You can stop it anytime.
Knowledge Builder uses compute resources while running. For cost efficiency, you may want to run it periodically rather than continuously, especially for large datasets.

​
Benefits for Analysis

When you create an analysis task, Knowledge Builder’s learning improves results:
  • Faster analysis — AI already knows your data structure
  • Better accuracy — Understanding of business semantics
  • Fewer errors — Awareness of data quality issues
  • Richer insights — Historical context and patterns