Analytics Engineer

Staging/Intermediate/Marts Layers

Por qué: Lets you restructure a project that grew organically, set naming and materialization rules, and stop marts from referencing raw sources directly.

Dimensional Modeling in Practice

Por qué: Lets you design conformed dimensions shared by several business areas, so sales and marketing report the same customer count rather than two competing ones.

Data Testing (Generic & Custom)

Por qué: Lets you decide which custom tests guard business-critical models and which are noise, keeping the suite fast enough that people do not skip it.

Data Quality Assertions

Por qué: Lets you set anomaly and distribution checks with sensible thresholds and an owner for each alert, so a quality incident is handled rather than muted.

Snowflake/BigQuery/Redshift

Por qué: Lets you weigh a warehouse migration or a multi-platform setup against pricing, concurrency and lock-in, and defend the recommendation to engineering and finance.

Cloud Cost Optimization for Analytics

Por qué: Lets you set up resource monitors and per-team cost attribution, so a runaway query is capped automatically and the monthly bill is explainable.

CI/CD for Data Models

Por qué: Lets you build slim CI with deferred state so a pull request tests only the changed models, giving fast feedback and blocking a breaking change before merge.

Data Lineage & Impact Analysis

Por qué: Lets you trace a wrong dashboard figure back to its source model and list every downstream asset affected before renaming a column.

Reverse ETL Patterns

Por qué: Lets you sync a modeled customer segment into the CRM safely, handling duplicates, failed syncs and field mapping, so sales acts on warehouse data.

Real-Time Analytics Modeling

Por qué: Lets you decide which metrics truly need near-real-time freshness and model them with micro-batch or streaming, leaving the rest in cheaper daily batch.