Duration: 6 months (extended from an initial 3-month internship)
During my internship within the Data team, I delivered two production-grade projects focused on:
- dbt transformation modelling
- BigQuery warehouse development
- Marketing data integration
- Pipeline orchestration and monitoring
- Documentation and BI enablement
All work was deployed to production and is actively used by the business.
- Documented the full data platform architecture in Figma
- Standardised visibility across all transformation pipelines
- Built dbt models powering analytics datasets
- Implemented data quality checks and tests
- Supported BI dashboard enablement
- Orchestrated and monitored DAGs in Airflow, ensuring smooth execution and timely reruns of failed tasks
| Tool | Purpose |
|---|---|
| Airbyte | Data ingestion from source systems |
| Airflow | Orchestration, scheduling, and DAG monitoring |
| BigQuery | Data warehouse |
| dbt | Data transformation and modelling |
| Metabase | Business Intelligence / reporting |
| DataHub | Data governance & cataloging |
This diagram shows how data is ingested via Airbyte, orchestrated with Airflow DAGs, transformed using dbt, and delivered to analytics dashboards.
Flow:
Source System (e.g., MySQL / Extracted CSVs) → Airbyte (Ingestion Pipelines) → BigQuery → dbt (staging → intermediate → marts) → Analytics Mart → Airflow DAGs for orchestration & monitoring → Metabase dashboards .
Notes:
- Source data simulates transactional or marketing platform data.
- Service accounts securely authenticated dbt and Airflow with BigQuery, following least-privilege principles.
- dbt layered modelling applied: staging → intermediate → marts. See an example dbt model here: stg_customers.sql
- Pipelines were orchestrated and monitored using Airflow, with Airbyte handling ingestion; failed runs were retried after corrections.
- DAG monitoring in Airflow included tracking task execution, retries, and dependencies to ensure timely delivery of analytics datasets.
This DAG is a public example illustrating pipeline orchestration. The screenshot shows a series of failed placeholder tasks (highlighted in red) labeled “empty 1”, “empty 2”, etc., which simulate tasks similar to those I monitored, troubleshooted, and reran during my internship to ensure reliable data pipelines.
-
Owned my own tickets in the team backlog, including pipelines for analysts
-
Contributed to daily team work and pipeline monitoring
-
Learned PowerShell scripting for automation tasks
-
Practiced branch management, pull requests, and code reviews:
- Incorporated feedback and reworked branches before merging to main
- Peer-reviewed teammates’ analytics engineering work once confident
-
Followed team governance: merges to main required sign-off by manager and a knowledgeable team member
- Production-style analytics engineering (BigQuery, dbt, Airbyte, Airflow)
- Pipeline orchestration and DAG monitoring
- Data modelling and star schema design
- Data quality testing
- Secure cloud access via GCP service accounts
- Git workflow & pull request management
- Peer review and collaborative teamwork
- BI enablement (Metabase dashboards)
- Documentation & visibility (Figma, DataHub)
This internship demonstrates:
- Real-world production data engineering experience
- Ability to implement warehouse-centric transformations
- Understanding of modern data stack architecture including orchestration with Airflow and ingestion with Airbyte
- Collaboration in a structured team with quality & governance processes