GISPulse — The Declarative Geospatial Engine

🗂️

16+ formats

GPKG, GeoJSON, Shapefile, GeoParquet, CSV+WKT, FlatGeobuf, WFS, GeoTIFF, PMTiles, PostGIS... no prior conversion needed.

Supported formats

The problem with current tools

What you deal with today

FME — 5,000 to 15,000 EUR/year per seat. Perpetual licenses deprecated in 2025. Total vendor lock-in.
QGIS Processing — Excellent for interactive use, but not headless, no API, no server mode or scheduling.
PostGIS alone — Powerful engine, but raw SQL to maintain, no native rule versioning, steep learning curve.
GeoPandas scripts — Flexible, but imperative: every change requires a developer, impossible to audit by a non-dev.

What GISPulse brings

Declarative JSON rules that your GIS team can read, modify and version without writing a single line of Python.
A portable engine — no database required. Run a processing job on any machine in 60 seconds.
The power of PostGIS when you need it — same rules, same file, just a different engine.
A REST API and Python SDK included — integrate GISPulse into your existing pipelines, ETLs, and apps.

FME: 5,000–15,000 EUR/year→GISPulse Community: free

A spatial pipeline in 5 lines

Identify parcels in flood zones, compute their area, and export statistics by municipality — all in a single rules file:

json

[
  {
    "name": "parcelles_a_risque",
    "capability": "spatial_join",
    "params": {
      "input": "parcelles.gpkg",
      "ref_layer": "zones_inondables.gpkg",
      "predicate": "intersects",
      "columns": ["niveau_risque"]
    }
  },
  {
    "name": "surface_par_commune",
    "capability": "spatial_aggregate",
    "params": {
      "input": "parcelles_a_risque",
      "ref_layer": "communes.gpkg",
      "predicate": "within",
      "agg": { "parcelle_id": "count", "surface_m2": "sum" }
    }
  }
]

bash

# Portable mode — no database required
gispulse run rules.json --engine duckdb

# API mode
curl -X POST http://localhost:8000/jobs \
  -d @rules.json

# SDK mode
from gispulse import GISPulse
gp = GISPulse()
result = gp.run("rules.json")

Same rules file, three execution modes: local CLI, REST API, Python SDK.

Three steps, zero friction

Import your data

GPKG, GeoJSON, Shapefile, GeoParquet, CSV, WFS... GISPulse reads everything, no prior conversion needed. The engine automatically adapts the schema.

Declare your rules

A JSON file describes the complete pipeline: capabilities to apply, parameters, execution order. Versionable under Git.

Export in the same format

The result is written in the original format. No database to manage, no conversion. Your files remain your files.

117 capabilities, ready to use

Vector — geometry

buffer, clip, union, dissolve, centroid, reproject
convex_hull, concave_hull, alpha_shape, voronoi_polygons, delaunay_triangulation
simplify, chaikin_smooth, offset_curve, line_merge, line_substring

Cross-layer analysis & overlay

spatial_join, intersects, spatial_aggregate, filter, nearest_neighbor
overlay_intersection, overlay_union, erase, merge_layers, classify_by_ring

Attributes & reshape

add_field, drop_field, rename_field, cast_field, attribute_join
case_when, coalesce_fields, lookup_table, pivot, unpivot
sort, deduplicate, random_sample, top_n

Classification & styling

classify (jenks / quantile / std_dev / pretty), classify_categorical
head_tail_breaks (Jiang 2013), normalize (log1p / minmax / zscore)
choropleth, bivariate_choropleth, continuous_ramp, graduated_size

Spatial statistics & clustering

morans_i, getis_ord_g, spatial_weights, kde_heatmap
cluster_kmeans, cluster_dbscan, cluster_hdbscan
grid_create, hexgrid_create

Topology & validation

topology_check, duplicate_geometry, attribute_validation, completeness_check
polygon_fix_gaps, polygon_fix_overlaps, polygon_remove_slivers
network_snap_endpoints, network_node_lines, network_remove_pseudo_nodes

3D pointcloud

pointcloud_load_las — LAS / LAZ to GeoDataFrame
pointcloud_filter_classification — ASPRS codes
pointcloud_zonal_height — building / canopy heights
pointcloud_grid_summary — Z statistics per grid cell

Raster & network (Pro)

zonal_stats, raster_clip, raster_reproject, ndvi, change_detection
shortest_path, isochrone, od_matrix, network_allocation, connectivity_check
postgis_sql — parameterized SQL query on PostGIS

Portable or persistent — same rules

📁

Portable mode

No database required. GISPulse spins up a temporary DuckDB engine in memory, executes your rules on your files, and writes the result in the original format.

GPKG in → DuckDB (memory) → GPKG out

Zero server installation
Works offline and in CI/CD
Output format = input format
Ideal for scripts, one-shots, batch ETL

📗

Persistent mode

Connect PostGIS. The same rules become real-time triggers, continuous pipelines, spatial materialized views.

PostGIS ↔ active triggers ↔ live results

Real-time spatial triggers
Continuous pipelines with scheduling
Multi-user with RBAC
Ideal for servers, APIs, dashboards

GISPulse vs the alternatives

Criteria	GISPulse	FME	QGIS Processing	PostGIS alone
Price	Free (AGPL)	5–15k EUR/year	Free	Free
Declarative	Native JSON	Visual GUI	GUI + scripts	Raw SQL
Portable mode	Embedded DuckDB	No	Desktop interface	No
Server mode	PostGIS / API	FME Flow ($$)	No	Yes
CLI / headless	Yes	Partial	No	psql only
REST API	Included	Paid	No	Not native
Versionable (Git)	Native (JSON)	XML export	Non-standard	SQL migrations
Python SDK	Included	Paid	PyQGIS scripts	psycopg2/SQLAlchemy
QGIS Plugin	Yes	Yes	Native	No
Cloud-native	S3, Docker, K8s	FME Cloud ($$)	No	Self-managed

Criteria

GISPulse

FME

QGIS Processing

PostGIS alone

Price

Free (AGPL)

5–15k EUR/year

Free

Declarative

Native JSON

Visual GUI

GUI + scripts

Raw SQL

Portable mode

Embedded DuckDB

Desktop interface

Server mode

PostGIS / API

FME Flow ($$)

Yes

CLI / headless

Yes

Partial

psql only

REST API

Included

Paid

Not native

Versionable (Git)

Native (JSON)

XML export

Non-standard

SQL migrations

Python SDK

Included

Paid

PyQGIS scripts

psycopg2/SQLAlchemy

QGIS Plugin

Yes

Native

Cloud-native

S3, Docker, K8s

FME Cloud ($$)

Self-managed

FME remains an excellent tool for complex graphical workflows. GISPulse targets teams that want to automate, version, and deploy their spatial processing in a modern pipeline. Detailed comparison →

GISPulseThe Declarative Geospatial Engine

Declarative

Multi-engine

16+ formats

Embeddable

Open Source AGPL-3.0

Enterprise-ready

The problem with current tools

What you deal with today

What GISPulse brings

A spatial pipeline in 5 lines

Three steps, zero friction

Import your data

Declare your rules

Export in the same format

117 capabilities, ready to use

Vector — geometry

Cross-layer analysis & overlay

Attributes & reshape

Classification & styling

Spatial statistics & clustering

Topology & validation

3D pointcloud

Raster & network (Pro)

Portable or persistent — same rules

Portable mode

Persistent mode

Pricing

GISPulse vs the alternatives

GISPulse v1.1.1 is available

GISPulseThe Declarative Geospatial Engine

Declarative

Multi-engine

16+ formats

Embeddable

Open Source AGPL-3.0

Enterprise-ready

The problem with current tools ​

What you deal with today

What GISPulse brings

A spatial pipeline in 5 lines ​

Three steps, zero friction ​

Import your data

Declare your rules

Export in the same format

117 capabilities, ready to use ​

Vector — geometry

Cross-layer analysis & overlay

Attributes & reshape

Classification & styling

Spatial statistics & clustering

Topology & validation

3D pointcloud

Raster & network (Pro)

Portable or persistent — same rules ​

Portable mode

Persistent mode

Pricing ​

GISPulse vs the alternatives ​

GISPulse v1.1.1 is available

The problem with current tools

A spatial pipeline in 5 lines

Three steps, zero friction

117 capabilities, ready to use

Portable or persistent — same rules

Pricing

GISPulse vs the alternatives