Geospatial Data Engineer

Company

Orcrist builds secure data intelligence software for defense, law enforcement, and enterprise teams. Our Sentinel platform combines data integration, AI-assisted analysis, and operational workflows. The GEOINT team is extending it with geospatial data services, remote-sensing capabilities, and a web-based common operational picture.
Role

Build the data foundation that makes satellite imagery and geospatial datasets discoverable, reliable, and ready for analysis. You'll use Python and Go to develop ingestion, processing, cataloging, and serving capabilities for data we acquire, retain, and use within our platform.

The work spans provider APIs, large raster deliveries, vector datasets, spatial databases, and object storage. You'll partner with Foundation and Platform engineers on infrastructure, and with data scientists and application engineers on the data products they need.
What you'll do

- Build provider integrations for imagery search, acquisition, download, and delivery tracking, handling authentication, rate limits, retries, and differences in vendor metadata.

- Ingest optical, multispectral, thermal, and SAR products from commercial and open sources. Our intended provider landscape includes Airbus Pléiades, Satellogic, SatVu, ICEYE, and Copernicus Sentinel missions.

- Develop raster ETL for metadata extraction, coordinate transformation, reprojection, resampling, mosaicking, COG creation, overviews, and quality checks.

- Build vector pipelines for geometry validation, schema normalization, spatial partitioning, indexing, and production of GeoParquet and PMTiles assets.

- Implement imagery cataloging with STAC/pgSTAC and spatial data services with PostGIS, preserving acquisition times, footprints, source metadata, processing history, and access restrictions.

- Design efficient storage and delivery paths using S3-compatible buckets, HTTP range requests, columnar formats, and appropriate caching and tiling strategies.

- Integrate processing with Sentinel's orchestration and service contracts. Make workflows resumable, safe to retry, observable, and capable of processing data larger than memory.

- Track freshness, completeness, lineage, and processing cost. Diagnose malformed deliveries, missing assets, inconsistent catalogs, and performance bottlenecks.

About you

- Several years of data or backend engineering experience, with substantial hands-on work on geospatial or Earth observation data.

- Strong Python skills and practical Go experience for services, integrations, or processing infrastructure.

- Experience with GDAL/Rasterio and vector tooling such as GeoPandas, Shapely, or equivalent libraries.

- Strong SQL and PostgreSQL/PostGIS knowledge, including spatial indexes, query planning, and efficient bulk loading.

- Experience handling formats such as GeoTIFF, GeoJSON, GeoPackage, and Shapefile, with a sound understanding of coordinate reference systems, nodata, geometry validity, and the effect of resampling on data quality.

- Experience with object storage and automated pipelines, including failure recovery, deduplication, metadata validation, and reproducible outputs.

- Comfortable packaging and operating software on Linux with containers, automated tests, and useful monitoring.

- Clear communication in English and an ability to agree data contracts with scientific and application teams. Eligible to work in Germany.

Nice‑to‑haves

- STAC/pgSTAC, PySTAC, TiTiler, Martin, or OGC data services.

- Apache Spark, Dask, xarray, DuckDB, Apache Arrow, or Apache Sedona for larger geospatial workloads; familiarity with multidimensional formats such as Zarr, NetCDF, and HDF5.

- Apache Iceberg or similar analytical table formats, including schema evolution, partitioning, and integration with data catalogs.

- Temporal, Kafka, Kubernetes, and operation in self-hosted or air-gapped environments.

- Provider delivery formats such as DIMAP, satellite acquisition APIs, or preparation of optical, therma