Data Engineer building batch and streaming pipelines on GCP and AWS, with a focus on data quality, reproducible infrastructure, and turning messy real-world data into decision-ready datasets.
- Pipelines & architecture — medallion / lakehouse patterns (Bronze → Silver → Gold), batch and streaming
- Transformations — SQL and dbt models that run the same locally (DuckDB) and in the cloud (BigQuery)
- Data quality — validation rules, confidence scoring, and tests that run in CI
- Infrastructure as code — Terraform, containers, and serverless (Cloud Run, AWS Lambda)
| Project | What it does | Stack |
|---|---|---|
| gcp-data-engineer-demand-pipeline | Batch pipeline that tracks data-engineering job demand in LATAM. Medallion architecture, dbt models, DuckDB locally and BigQuery + Cloud Run on GCP, plus an MCP server for LLM querying. | Python · dbt · DuckDB · BigQuery · Cloud Run · Terraform |
| SmartCity | Real-time smart-city streaming pipeline for vehicle, GPS, weather and traffic data, stored as Parquet. | Kafka · Spark · S3 · Docker |
| ETL-Pipeline-Environmental-Crimes | ETL from a public API to PostgreSQL and S3, with structured logging, unit tests and serverless ingestion. | Python · pandas · AWS Lambda · S3 · PostgreSQL |
| NLP-for-IPCC-Climate-Reports | Web scraping, PDF processing and NLP to extract and analyze IPCC climate reports. | Python · spaCy · pandas |
| AuriData | Compact, containerized ETL system for historical loads with strict validation. | Python · Docker |
| IRCA-Analysis-System | CLI that computes water-quality risk (IRCA) from geolocated records. | Rust · DuckDB |
- Organizer at Google Developers Group (GDG)
- Regular participant in data-centric hackathons, using my environmental engineering background to extract insights from environmental and climate data

