// data engineer · 3 years · AWS
I've spent 3 years designing, building and optimizing ETL/ELT pipelines with Python, SQL, PySpark and AWS (Glue, Redshift, S3, EMR). I'm part of an AI-focused team, building the data infrastructure behind model training, inference and analytics. I'm now expanding into real-time streaming with Apache Kafka.
AI-focused engineering team. I own the data layer that powers AI/ML inference and analytics.
Batch data pipelines and data warehousing for BI and MIS reporting.
End-to-end ETL workflows that fetch data from external REST APIs and turn it into structured, analytics-ready datasets. They include cleaning, validation and normalization logic, plus logging, error handling and retries. Outbound integrations deliver model-ready data to downstream AI/ML training and inference services.
A scalable batch pipeline in AWS Glue (PySpark) that processes raw transactional data from S3. It applies joins, aggregations and window functions, loads Redshift, and runs post-load stored procedures automatically. Validation, schema enforcement and partitioning are built in, and the full automation removed manual reporting work.
A simple retrieval-augmented generation app that runs locally, built to learn the RAG pattern end to end: chunking documents, embedding, retrieval, and answer generation through a free-tier LLM API.
A streaming pipeline for banking-style transaction events on AWS, built as part of my move into real-time data with Apache Kafka.
Open to conversations about data engineering and AI data infrastructure.