Skip to content

Cloud, Platform, and Community Foundations

Geospatial software engineering is shaped by both scientific foundations and the platforms engineers actually use in production. This book treats AWS, Google Cloud, Microsoft Azure, Esri, Mapbox, CARTO, QGIS, and community forums as practical learning ecosystems.

Platform Pattern

Most production geospatial platforms can be understood as this pipeline:

source data -> storage -> index/catalog -> compute -> API -> map/model/user workflow

The same pattern appears in different forms:

  • AWS: object storage, serverless/event services, databases, and Amazon Location Service for maps, places, routing, tracking, and geofencing.
  • Google Cloud: BigQuery GEOGRAPHY, Earth Engine, Cloud Storage, notebooks, and visualization tools.
  • Microsoft Azure: Azure Maps, Azure storage, Azure Functions, Azure Kubernetes Service, Azure DevOps, and analytics services.
  • Esri: ArcGIS Enterprise, ArcGIS Online, hosted feature layers, image services, vector tile layers, routing, geocoding, GeoEnrichment, and developer SDKs.
  • Mapbox: map rendering, vector tiles, search/geocoding, navigation, tiling services, data products, and developer SDKs.
  • CARTO: cloud-native spatial analytics and location intelligence running on cloud data warehouses.
  • QGIS: desktop analysis, data inspection, processing models, cartography, plugins, and training workflows.

Core Equations for Cloud Geospatial Engineering

Cost Model

total_cost =
  storage_cost
  + compute_cost
  + data_transfer_cost
  + request_cost
  + licensing_or_subscription_cost

For a pipeline:

pipeline_cost_per_run =
  input_bytes * storage_read_rate
  + compute_seconds * compute_rate
  + output_bytes * storage_write_rate
  + api_calls * request_rate

Latency Budget

total_latency =
  network_latency
  + authentication_latency
  + query_latency
  + processing_latency
  + serialization_latency
  + rendering_latency

Use this when designing map APIs, tile services, GeoRAG retrieval, routing systems, and dashboards.

Throughput

throughput = processed_records / elapsed_time

For spatial pipelines:

features_per_second = feature_count / processing_seconds
pixels_per_second = pixel_count / processing_seconds
tiles_per_second = tile_count / rendering_seconds

Cache Hit Rate

cache_hit_rate = cache_hits / (cache_hits + cache_misses)

Tile services, geocoding, routing, and GeoRAG retrieval all benefit from caching, but cached outputs need invalidation and provenance.

Tile Cardinality

At zoom level z, a Web Mercator tile pyramid has:

tile_count(z) = 4^z

For a bounded region:

regional_tile_count = count(tiles intersecting region at zoom z)

This is why global high-zoom tiles are expensive and why vector tiles, simplification, and caching matter.

Warehouse Spatial Join Cost

Conceptually:

join_cost ~= candidate_pairs_after_index_or_partitioning * exact_predicate_cost

Good design reduces candidate pairs before running exact geometry predicates.

GeoRAG Retrieval Budget

retrieval_latency =
  spatial_filter_latency
  + vector_search_latency
  + keyword_search_latency
  + reranking_latency
  + citation_packaging_latency

GeoRAG quality depends on both retrieval score and spatial/temporal eligibility:

eligible_document =
  spatial_relation(document.geometry, query.geometry)
  and temporal_relation(document.time, query.time)
  and permission_allows_access(document, user)

Platform-Specific Learning Notes

AWS

Amazon Location Service provides managed location functionality including maps, places, routes, tracking, and geofencing. In a geospatial architecture, it often supplies application-facing location services while other AWS services handle storage, events, analytics, and deployment.

Google Cloud

BigQuery geospatial analytics uses the GEOGRAPHY data type and GoogleSQL geography functions. It is useful when spatial analysis needs to run close to large analytical tables. Google Earth Engine and BigQuery-connected workflows are important references for Earth observation and large-scale analysis.

Microsoft Azure

Azure Maps provides mapping and geospatial services for maps, geocoding, search, routing, traffic, geolocation, time zones, weather, and spatial operations. Azure is also important for DevOps, virtual training environments, storage, identity, and enterprise deployment.

Esri

Esri's developer platform includes mapping and location services, portal and data services, spatial analysis services, raster analysis, offline workflows, security/authentication, and SDKs. Esri remains central in enterprise GIS, utilities, local government, and field operations.

Mapbox

Mapbox provides maps, search/geocoding, navigation, vector tiles, tiling services, mobile SDKs, GL rendering, and data products. It is especially useful for custom web/mobile experiences, high-performance rendering, and developer-centered location applications.

CARTO

CARTO focuses on cloud-native location intelligence and spatial analytics on top of cloud data warehouses. It is useful for analytics teams that want spatial workflows close to BigQuery, Snowflake, Redshift, Databricks, PostgreSQL, and similar platforms.

QGIS

QGIS is a foundational learning and inspection tool. The QGIS Training Manual covers vector analysis, raster analysis, network analysis, spatial statistics, PostGIS, processing models, interpolation, hydrology, and many reproducible exercises.

Spatial Community Slack

The Spatial Community is a Slack-based community with more than 10,000 geospatial enthusiasts. It is valuable for practical wisdom: debugging PostGIS queries, discussing Python libraries, learning GeoServer patterns, reviewing architecture choices, and discovering how working practitioners solve messy real-world problems.

Sources

  • Amazon Location Service Documentation. Amazon Web Services, https://aws.amazon.com/documentation-overview/location/. Accessed 9 May 2026.
  • "Introduction to Geospatial Analytics." Google Cloud BigQuery, https://cloud.google.com/bigquery/docs/geospatial-intro. Accessed 9 May 2026.
  • "Azure Maps Documentation." Microsoft Learn, https://learn.microsoft.com/en-us/azure/azure-maps/. Accessed 9 May 2026.
  • "Developer Documentation." Esri Developer, https://developers.arcgis.com/documentation/. Accessed 9 May 2026.
  • "Mapbox Documentation." Mapbox, https://docs.mapbox.com/. Accessed 9 May 2026.
  • "What Is CARTO?" CARTO Documentation, https://docs.carto.com/getting-started/carto-in-a-nutshell. Accessed 9 May 2026.
  • "QGIS Training Manual." QGIS Documentation, https://docs.qgis.org/latest/en/docs/training_manual/index.html. Accessed 9 May 2026.
  • "The Spatial Community." The Spatial Community, https://thespatialcommunity.org/. Accessed 9 May 2026.

GeoInformatica Consulting logo