Amelia – SQL, Airflow, Python

It's a data and conversational automation platform designed to ingest data daily from an external API, process both historical and incremental datasets, and expose insights through a WhatsApp bot powered by an LLM.

• Designed the end-to-end data architecture.
• Implemented scalable ETL pipelines with Apache Airflow, processing a 24-month historical backfill and daily incremental updates.
• Modeled data to support both analytical and conversational queries.
• Integrated a WhatsApp chatbot with an LLM via n8n, ensuring reliability, observability, and cost-efficient MVP operation.

It's a custom Trino connector for a proprietary big data and AI-oriented database, enabling distributed SQL query execution and metadata federation across large-scale datasets.

• Designed and implemented core connector components for schema discovery, query planning, and predicate pushdown.
• Optimized performance for handling large datasets.
• Ensured full compatibility with the Trino SPI.
• Collaborated with stakeholders to align technical solutions with production requirements.

It's a cloud-native data platform on AWS, supporting both batch and analytical workloads. The platform enabled data ingestion, transformation, governance, and visualization, providing high-quality datasets and insights for business decision-making.

• Built and maintained scalable data pipelines.
• Implemented data modeling and governance strategies.
• Optimized analytical queries using Iceberg and Athena.
• Delivered interactive data visualizations to support business stakeholders.

It's a robust data platform handling both real-time streaming and batch processing workloads.

• Implemented and optimized streaming and batch data pipelines.
• Improved data reliability, scalability, and performance.
• Managed infrastructure as code using Terraform.
• Collaborated with cross-functional teams to deliver robust data solutions.

It's a large-scale enterprise data platform following data mesh principles. The platform enables domain-oriented data ownership, scalable data products, and both batch and streaming analytics across multiple business units.

• Contributed to the design and implementation of a data mesh architecture.
• Built scalable and maintainable data pipelines.
• Enhanced observability and system reliability.
• Supported multiple teams in delivering high-quality data products.