This talk presents a case study of Inteligeo, an adaptation of GeoNode for Brazilian government agencies. We'll share our journey in taming a 170TB raster dataset and other adventures in customizing open-source geospatial tools. We outline the experience of four agencies: The National Water Agency (ANA), the Center for the Surveillance of the Amazon (CENSIPAM), the Chico Mendes Institute for Biodiversity Conservation (ICMBIO), and the Federal Police (PF).

The project, which began in 2009 using proprietary software, transitioned to open-source in 2017. To address limitations and support multiple agencies, we ported Inteligeo version 4 functionality to GeoNode, creating Inteligeo 5 in 2022. Each agency has a unique perspective: PF was involved from the start, CENSIPAM deployed for internal use, ANA for external use, and ICMBIO is integrating it into their processes.

Why customize at all? Why not just use the software as it is? Why not develop it from scratch? There are several reasons why one should (and shouldn't) customize, and then there are several ways how to do it right once you commit to it. We share our experiences: the good and the bad, and the lessons learned, when customizing GeoNode for the Brazilian government.

Why customize?
- Get shiny new functionality!
- Jump-start development
- Optimize processes with tailored deployments and workflows
- Integrate with existing systems and infrastructure (authentication, 170TB raster storage, Brasil Mais imagery)
- Comply with internal and government standards

Challenges:
- Non-standard deployment
- Training and documentation
- Balancing customization with community support
- Syncing with upstream changes
- Managing a huge codebase with extra stuff that you don't need

How to do it right (aspirational):
- Seek sponsorships (SGD/MGI, JICA, FINEP, INTERPOL)
- Selfless and selfish reasons to contribute back to the community
- Have a clear strategy for upstream syncing
- Keep it simple. Minimize customization to essential features
- Design independent, standalone components
- Engage upstream developers when possible

Our experience is particularly relevant to the Amazon region, as the tool directly supports the Federal Police surveillance and conservation efforts in the area and is being integrated by the agencies of the other speakers.

We welcome feedback and collaboration ideas from the FOSS4G community during the Q&A session!

This talk presents case studies of deploying GeoNode, an open-source geospatial content management system, in production environments within two Brazilian government agencies: the Geological Survey of Brazil (SGB) and the Brazilian Federal Police (PF). We'll explore how these agencies have successfully implemented and customized GeoNode to meet their specific needs, addressing common challenges in large-scale FOSS4G deployments.

Key points we'll cover:

1. SGB's approach:
   - Developing a Helm chart for automated GeoNode 4 installation on Red Hat OpenShift
   - Addressing security requirements like rootless execution and random UID support
   - Implementing autoscaling for most components based on CPU and memory utilization
   - Exploring cluster implementation of GeoServer for improved scalability

2. PF's customizations:
   - Creating a dedicated "inteligeo-deploy" repository for enhanced deployment features
   - Implementing centralized configuration and logging
   - Improving security by separating credentials and using Podman instead of Docker
   - Integrating with internal systems and scheduling data updates

We'll discuss the challenges faced, solutions implemented, and lessons learned from both approaches. These case studies demonstrate that FOSS4G solutions like GeoNode are ready for production use in government agencies, providing flexibility, scalability, and security.

By sharing our experiences, we aim to help other organizations successfully deploy GeoNode and other FOSS4G solutions in production environments. We welcome questions and discussions on best practices for large-scale FOSS4G implementations.

This talk explores how the Brazilian Federal Police leverages FOSS4G to integrate and visualize data from Brasil MAIS, a multi-million dollar environmental monitoring program, using our custom-built Inteligeo platform.

Brasil MAIS monitors over 2.9 million km² weekly - about 34% of Brazil's vast 8.5 million km² territory. To manage the resulting massive amounts of satellite imagery and environmental change alerts, we developed Inteligeo, a GeoNode-based solution currently in production at the Federal Police and in various stages of adoption across multiple government agencies.

We'll discuss:
1. Building Inteligeo on FOSS4G principles (GeoNode, PostGIS, GDAL, GeoServer, FastAPI) to enable seamless data integration across government agencies.
2. Implementing user-friendly tools for non-specialists and integrating with Gov.br authentication.
3. Using FastAPI for complex integrations, such as streamlining access to the XYZ data provider for Brasil Mais.
4. Technical challenges in scaling Brasil MAIS to handle 46 million monthly tile views and serve 500+ public institutions and 100,000+ users.
5. The crucial sponsorship of the Ministry of Management and Innovation in Public Services (MGI) in developing Inteligeo 5 since 2021, and their plans for its independent use.

Brasil MAIS has already demonstrated significant impact, contributing to over R$16 billion ($2.8 billion USD) in environmental crime-related fines and asset freezes. While Inteligeo is still ramping up adoption, it aims to compound this impact by enhancing the efficiency of data utilization. We anticipate measuring Inteligeo-specific results in the coming year, showcasing how FOSS4G solutions can amplify the effectiveness of large-scale environmental monitoring programs.

We'll conclude by examining how this integration serves as a model for similar global initiatives in environmental management and welcome input from the FOSS4G community on further improvements.