Hydrosat
- Lead architecture and development of satellite imagery processing pipelines, from ingestion and scientific processing through quality control and delivery. Translate requirements from Data Science and Product into service boundaries and engineering milestones.
- Design distributed processing on AWS with Airflow, Kubernetes, SQS, and Postgres, with workload scaling, retries, failure isolation, and reprocessing to support reliable pipeline operation.
- Package scientific models as versioned Python libraries, separating model logic from orchestration and infrastructure to improve testability and reuse across workflows.
- Define pipeline observability and data-quality checks using processing metadata, correlation identifiers, metrics, and lineage to trace data across processing stages.
Technologies: Python, AWS, Airflow, Kubernetes, SQS, Postgres, Docker, Terraform/OpenTofu, GDAL, Rasterio