Geospatial software engineering

From raw survey
to running system.

We build custom geospatial software for survey, mapping, utility and environmental companies. Platforms, pipelines and 3D, from the data model to production.

01 · Raw data

It starts as measurements.

A LiDAR survey is millions of points, each one a measurement of the ground. This is five kilometres of the Grand Canyon.

02 · Information

We turn it into information.

Surfaces, contours and layers, computed by pipelines that are checked, repeatable and never made by hand.

03 · Systems

And into systems people use.

Maps, platforms and services that your team and your clients open every day.

Scroll

Point cloud: USGS 3DEP, Grand Canyon National Park, 2019

What we solve

For the problems off-the-shelf GIS leaves behind.

01

You have the data. You want a product.

Survey and mapping firms sit on years of deliverables. We build the platform, portal or service that lets you sell what you already produce.

02

Your pipeline runs on one person.

Ingestion, processing and publishing that depend on someone at a desktop become automated pipelines that either finish or fail where you can see it.

03

Your 3D ends up as files.

Point clouds, meshes and city models streamed in the browser, measured and shared, instead of shipped on a drive.

04

It slows down when it matters.

Tiles, APIs and spatial databases tuned for peak load, served from the edge and watched from the outside.

05

Nobody knows how it all fits anymore.

We take over inherited geospatial platforms: document them, monitor them, run them, then make them better.

06

You want AI that understands maps.

Plain-language questions over spatial data, classification and extraction, built on your data rather than a demo.

Selected work

Systems in production, for teams on four continents.

Client names stay with our clients. These are the systems and what they changed.

01A Swedish forestry-tech company

Drone flights become finished map layers.

Field teams submit a flight. A fleet of processing machines turns the photos into orthomosaics and elevation models and publishes them as layers the customer opens the next morning. Every mission, job and layer is tracked, and nobody touches a terminal.

PhotogrammetryDistributed processingWeb mapping
  • Our largest system
  • Built and operated by us
02An Australian weather company owned by a media group

An inherited platform, now documented, monitored and run.

After the platform changed owners and clouds, how it fit together lived in people's heads. We documented it from the code up, put it under monitoring from the outside, and now run its development, operations and incidents.

KubernetesEdge deliveryMonitoring
  • 63 external health checks
  • Moved between clouds
03The same weather company

The weather map its partners embed.

Eight days of forecast from seven models, radar replayed every five minutes, satellite, observations and warnings, served from the edge to the company's own site and to its partners.

Vector tilesForecast dataMapLibre
  • 7 forecast models
  • Radar every 5 minutes
04A US hydrographic survey firm

A survey goes in. 3D and 2D products come out.

Multibeam sonar and LiDAR become streamable point clouds, elevation models, hillshades and contours, in a viewer where the 3D scene and the 2D map move together. Before, every one of those assets was made by hand.

LiDARBathymetry3D + 2D
  • ~70 survey sites
  • Clouds up to 375M points
  • In production since May 2026
05A Brazilian aerial mapping company

Photorealistic 3D cities, in the browser.

Oblique aerial surveys stream as 3D Tiles. People measure in 3D, browse the oblique imagery and fly through the city. It began as a prototype we built and became the company's own product.

3D TilesCesiumOblique imagery
  • Prototype to product
06A point-forecast API

Peak-hour crashes, ended.

A point forecast cut from gridded output took 10 to 15 seconds and fell over at peak. A gateway and an on-demand store took server errors to zero, without rewriting the backend.

APIsCachingGridded data
Read the story
  • Server errors to zero

Clients

Teams we've built for.

Skyforest

Skyforest

Forestry and environmental monitoring

Jane's Weather

Jane's Weather

Weather forecasting

ACM

ACM

Media and agricultural weather

Underwater Acoustics International

Underwater Acoustics International

Hydrographic survey

Topocart

Topocart

Aerial mapping and cadastre

TeroAI

TeroAI

AI-native geospatial intelligence

Gestão Engenharia

Gestão Engenharia

Engineering and project management

Cloudmatize

Cloudmatize

Cloud and data engineering

How we work

Start small, or start big.

01

A focused review

A look at one platform, pipeline or architecture, with a written plan of what to fix first. Weeks, not months.

02

A build

From the data model to production: we scope it with you, build it, and hand you a system your team can run.

03

Ongoing engineering

We run and improve what you already have: development, operations, monitoring and on-call.

Remote, worldwide. In English or Portuguese.

Who we are

A geospatial engineering company in Brazil.

Our engineers come from surveying and cartography as much as from software. They understand the data, its coordinate systems, its precision and the deliverables your clients expect, as well as the code that serves it.

Victor Marotta

Founder & Lead Engineer

Survey and cartographic engineer, MSc in civil engineering on GIS and spatial databases.

LinkedIn

Julia Gomes de Menezes

Managing Director

Leads business operations, strategy and growth: finance, contracts and partnerships.

LinkedIn

Contact

Tell us what you're working on.

A few lines are enough. We reply within one business day.

Or write to us directly

[email protected] LinkedIn

Brazil · working worldwide