dbt Models / Transformation Layer
Transformation layer built in dbt on BigQuery. Models handle
data cleaning, normalization of inconsistent labels, and
aggregation logic for energy consumption metrics by household
segment, time of day, and weather conditions.
1
7
GitHub README / Documentation
Full project documentation covering architecture decisions,
technology choices, setup instructions, and pipeline flow.
Written to be reproducible by any engineer — no tribal knowledge
required. Reflects my approach to building infrastructure that
outlasts the person who built it.
1
9
Client-facing dashboard built in Looker Studio on top of the
BigQuery data model. Includes KPI scorecards, time-series energy
consumption analysis, hourly demand heatmap by socioeconomic
segment, and interactive filters. Designed to communicate
insights to non-technical stakeholders.
1
19
Pipeline Architecture
End-to-end data pipeline built on GCP for the London Smart Meters
dataset. Designed the full architecture: ingestion with dlt,
orchestration with Kestra, transformation with dbt and Spark,
and storage in BigQuery with partitioned and clustered tables
for query optimization.