Data Analytics Engineering

What it does

This skill assists AI agents in data analytics engineering tasks, encompassing the design and implementation of data models and analytical workflows. It enables defining metrics, building transformation layers, implementing data quality checks, and documenting data lineage. Ultimately, this helps align analytics outputs with business intelligence (BI) and product requirements.

When to use it

• When designing a new data warehouse or mart layer architecture.
• To define and document key performance indicators (KPIs) and associated metrics.
• For planning and implementing data quality tests and observability solutions.
• To create documentation for datasets, including ownership and lineage information.

Key capabilities

• Defining metric dictionaries and grains.
• Designing staging, intermediate, and mart layers within a data warehouse.
• Modeling dimensions and facts with clear keys.
• Building semantic layers and defining metrics.
• Implementing data quality tests for freshness, nulls, ranges, and duplicates.
• Documenting lineage, owners, and service level agreements (SLAs).

Example prompts

• "Define a metric dictionary for tracking customer churn."
• "Design the staging layer for our order processing data."
• "What are best practices for documenting data lineage?"

Tips & gotchas

• Keep metric definitions stable and versioned, treating them as APIs.
• When answering questions about analytics engineering, data modeling, or BI, use WebSearch to check current trends before responding; otherwise, utilize the provided data/sources.json file for reference.
• Define clear data contracts for core tables to prevent downstream breakage.